Wearable sensors have become ubiquitous thanks to a variety of health tracking features. The resulting continuous and longitudinal measurements from everyday life generate large volumes of data. However, making sense of these observations for scientific and actionable insights is non-trivial. Inspired by the empirical success of generative modeling, where large neural networks learn powerful representations from vast amounts of text, image, video, or audio data, we investigate the scaling properties of wearable sensor foundation models across compute, data, and model size. Using a dataset of up to 40 million hours of in-situ heart rate, heart rate variability, accelerometer, electrodermal activity, skin temperature, and altimeter per-minute data from over 165,000 people, we create LSM, a multimodal foundation model built on the largest wearable-signals dataset with the most extensive range of sensor modalities to date. Our results establish the scaling laws of LSM for tasks such as imputation, interpolation and extrapolation across both time and sensor modalities. Moreover, we highlight how LSM enables sample-efficient downstream learning for tasks including exercise and activity recognition.

We present SensorLM, a family of sensor-language foundation models that enable wearable sensor data understanding with natural language. Despite its pervasive nature, aligning and interpreting sensor data with language remains challenging due to the lack of paired, richly annotated sensor-text descriptions in uncurated, real-world wearable data. We introduce a hierarchical caption generation pipeline designed to capture statistical, structural, and semantic information from sensor data. This approach enabled the curation of the largest sensor-language dataset to date, comprising over 59.7 million hours of data from more than 103,000 people. Furthermore, SensorLM extends prominent multimodal pretraining architectures (e.g., CLIP, CoCa) and recovers them as specific variants within a generic architecture. Extensive experiments on real-world tasks in human activity analysis and healthcare verify the superior performance of SensorLM over state-of-the-art in zero-shot recognition, few-shot learning, and cross-modal retrieval. SensorLM also demonstrates intriguing capabilities including scaling behaviors, label efficiency, sensor captioning, and zero-shot generalization to unseen tasks. Code is available at https://github.com/Google-Health/consumer-health-research/tree/main/sensorlm.

How should a machine intelligence perform unsupervised structure discovery over streams of sensory input? One approach to this problem is to cast it as an apperception task. Here, the task is to construct an explicit interpretable theory that both explains the sensory sequence and also satisfies a set of unity conditions, designed to ensure that the constituents of the theory are connected in a relational structure. However, the original formulation of the apperception task had one fundamental limitation: it assumed the raw sensory input had already been parsed using a set of discrete categories, so that all the system had to do was receive this already-digested symbolic input, and make sense of it. But what if we don't have access to pre-parsed input? What if our sensory sequence is raw unprocessed information? The central contribution of this paper is a neuro-symbolic framework for distilling interpretable theories out of streams of raw, unprocessed sensory experience. First, we extend the definition of the apperception task to include ambiguous (but still symbolic) input: sequences of sets of disjunctions. Next, we use a neural network to map raw sensory input to disjunctive input. Our binary neural network is encoded as a logic program, so the weights of the network and the rules of the theory can be solved jointly as a single SAT problem. This way, we are able to jointly learn how to perceive (mapping raw sensory information to concepts) and apperceive (combining concepts into declarative rules).

How should a machine intelligence perform unsupervised structure discovery over streams of sensory input? One approach to this problem is to cast it as an apperception task [1]. Here, the task is to construct an explicit interpretable theory that both explains the sensory sequence and also satisfies a set of unity conditions, designed to ensure that the constituents of the theory are connected in a relational structure. However, the original formulation of the apperception task had one fundamental limitation: it assumed the raw sensory input had already been parsed using a set of discrete categories, so that all the system had to do was receive this already-digested symbolic input, and make sense of it. But what if we don't have access to pre-parsed input? What if our sensory sequence is raw unprocessed information? The central contribution of this paper is a neuro-symbolic framework for distilling interpretable theories out of streams of raw, unprocessed sensory experience. First, we extend the definition of the apperception task to include ambiguous (but still symbolic) input: sequences of sets of disjunctions. Next, we use a neural network to map raw sensory input to disjunctive input. Our binary neural network is encoded as a logic program, so the weights of the network and the rules of the theory can be solved jointly as a single SAT problem. This way, we are able to jointly learn how to perceive (mapping raw sensory information to concepts) and apperceive (combining concepts into declarative rules).

This paper attempts to answer a central question in unsupervised learning: what does it mean to “make sense” of a sensory sequence? In our formalization, making sense involves constructing a symbolic causal theory that both explains the sensory sequence and also satisfies a set of unity conditions. The unity conditions insist that the constituents of the causal theory – objects, properties, and laws – must be integrated into a coherent whole. On our account, making sense of sensory input is a type of program synthesis, but it is unsupervised program synthesis. Our second contribution is a computer implementation, the Apperception Engine, that was designed to satisfy the above requirements. Our system is able to produce interpretable human-readable causal theories from very small amounts of data, because of the strong inductive bias provided by the unity conditions. A causal theory produced by our system is able to predict future sensor readings, as well as retrodict earlier readings, and impute (fill in the blanks of) missing sensory readings, in any combination. In fact, it is able to do all three tasks simultaneously. We tested the engine in a diverse variety of domains, including cellular automata, rhythms and simple nursery tunes, multi-modal binding problems, occlusion tasks, and sequence induction intelligence tests. In each domain, we test our engine's ability to predict future sensor values, retrodict earlier sensor values, and impute missing sensory data. The Apperception Engine performs well in all these domains, significantly out-performing neural net baselines. We note in particular that in the sequence induction intelligence tests, our system achieved human-level performance. This is notable because our system is not a bespoke system designed specifically to solve intelligence tests, but a general-purpose system that was designed to make sense of any sensory sequence.

Recent research has made the surprising finding that state-of-the-art deep learning models sometimes fail to generalize to small variations of the input. Adversarial training has been shown to be an effective approach to overcome this problem. However, its application has been limited to enforcing invariance to analytically defined transformations like lp-norm bounded perturbations. Such perturbations do not necessarily cover plausible real-world variations that preserve the semantics of the input (such as a change in lighting conditions). In this paper, we propose a novel approach to express and formalize robustness to these kinds of real-world transformations of the input. The two key ideas underlying our formulation are (1) leveraging disentangled representations of the input to define different factors of variations, and (2) generating new input images by adversarially composing the representations of different images. We use a StyleGAN model to demonstrate the efficacy of this framework. Specifically, we leverage the disentangled latent representations computed by a StyleGAN model to generate perturbations of an image that are similar to real-world variations (like adding make-up, or changing the skin-tone of a person) and train models to be invariant to these perturbations. Extensive experiments show that our method improves generalization and reduces the effect of spurious correlations (reducing the error rate of a "smile" detector by 21% for example).

In this paper we propose to augment a modern neuralnetwork architecture with an attention model inspired by human perception. Specifically, we adversarially train and analyze a neural model incorporating a human inspired, visual attention component that is guided by a recurrent top-down sequential process. Our experimental evaluation uncovers several notable findings about the robustness and behavior of this new model. First, introducing attention to the model significantly improves adversarial robustness resulting in state-of-the-art ImageNet accuracies under a wide range of random targeted attack strengths. Second, we show that by varying the number of attention steps (glances/fixations) for which the model is unrolled, we are able to make its defense capabilities stronger, even in light of stronger attacks - resulting in a “computational race” between the attacker and the defender. Finally, we show that some of the adversarial examples generated by attacking our model are quite different from conventional adversarial examples - they contain global, salient and spatially coherent structures coming from the target class that would be recognizable even to a human, and work by distracting the attention of the model away from the main object in the original image.

Does a Variational AutoEncoder (VAE) consistently encode typical samples generated from its decoder? This paper shows that the perhaps surprising answer to this question is `No'; a (nominally trained) VAE does not necessarily amortize inference for typical samples that it is capable of generating. We study the implications of this behaviour on the learned representations and also the consequences of fixing it by introducing a notion of self consistency. Our approach hinges on an alternative construction of the variational approximation distribution to the true posterior of an extended VAE model with a Markov chain alternating between the encoder and the decoder. The method can be used to train a VAE model from scratch or given an already trained VAE, it can be run as a post processing step in an entirely self supervised way without access to the original training data. Our experimental analysis reveals that encoders trained with our self-consistency approach lead to representations that are robust (insensitive) to perturbations in the input introduced by adversarial attacks. We provide experimental results on the ColorMnist and CelebA benchmark datasets that quantify the properties of the learned representations and compare the approach with a baseline that is specifically trained for the desired property.

Convex relaxations have emerged as a promising approach for verifying properties of neural networks, but widely used using Linear Programming (LP) relaxations only provide meaningful certificates when networks are specifically trained to facilitate verification. This precludes many important applications which involve \emph{verification-agnostic} networks that are not trained specifically to promote verifiability. On the other hand, semidefinite programming (SDP) relaxations have shown success on verification-agnostic networks, such as adversarially trained image classifiers without additional regularization, but do not currently scale beyond small networks due to poor time and space asymptotics. In this work, we propose a first-order dual SDP algorithm that provides (1) any-time bounds (2) requires memory only linear in the total number of network activations and (3) has per-iteration complexity that scales linearly with the complexity of a forward/backward pass through the network. By exploiting iterative eigenvector methods, we express all solver operations in terms of forward and backward passes through the network, enabling efficient use of hardware optimized for deep learning. This allows us to dramatically improve the magnitude of $\ell_\infty$ perturbations for which we can verify robustness verification-agnostic networks ($1\% \to 88\%$ on MNIST, $6\%\to 40\%$ on CIFAR-10). We also demonstrate tight verification for a quadratic stability specification for the decoder of a variational autoencoder.

The instability of Generative Adversarial Network (GAN) training has frequently been attributed to gradient descent. Consequently, recent methods have aimed to tailor the models and training procedures to stabilise the discrete updates. In contrast, we study the continuous-time dynamics induced by GAN training. Both theory and toy experiments suggest that these dynamics are in fact surprisingly stable. From this perspective, we hypothesise that instabilities in training GANs arise from the integration error in discretising the continuous dynamics. We experimentally verify that well-known ODE solvers (such as Runge-Kutta) can stabilise training - when combined with a regulariser that controls the integration error. Our approach represents a radical departure from previous methods which typically use adaptive optimisation and stabilisation techniques that constrain the functional space (e.g. Spectral Normalisation). Evaluation on CIFAR-10 and ImageNet shows that our method outperforms several strong baselines, demonstrating its efficacy.

A fundamental component of neural network verification is the computation of bounds on the values their outputs can take. Previous methods have either used off-the-shelf solvers, discarding the problem structure, or relaxed the problem even further, making the bounds unnecessarily loose. We propose a novel approach based on Lagrangian Decomposition. Our formulation admits an efficient supergradient ascent algorithm, as well as an improved proximal algorithm. Both the algorithms offer three advantages: (i) they yield bounds that are provably at least as tight as previous dual algorithms relying on Lagrangian relaxations; (ii) they are based on operations analogous to forward/backward pass of neural networks layers and are therefore easily parallelizable, amenable to GPU implementation and able to take advantage of the convolutional structure of problems; and (iii) they allow for anytime stopping while still providing valid bounds. Empirically, we show that we obtain bounds comparable with off-the-shelf solvers in a fraction of their running time, and obtain tighter bounds in the same time as previous dual algorithms. This results in an overall speed-up when employing the bounds for formal verification. Code for our algorithms is available at https://github.com/oval-group/decomposition-plnn-bounds.

The success of Deep Learning and its potential use in many safety-critical applicationshas motivated research on formal verification of Neural Network (NN) models. In thiscontext, verification involves proving or disproving that an NN model satisfies certaininput-output properties. Despite the reputation of learned NN models as black boxes,and the theoretical hardness of proving useful properties about them, researchers havebeen successful in verifying some classes of models by exploiting their piecewise linearstructure and taking insights from formal methods such as Satisifiability Modulo Theory.However, these methods are still far from scaling to realistic neural networks. To facilitateprogress on this crucial area, we exploit the Mixed Integer Linear Programming (MIP) formulation of verification to propose a family of algorithms based on Branch-and-Bound (BaB). We show that our family contains previous verification methods as special cases.With the help of the BaB framework, we make three key contributions. Firstly, we identifynew methods that combine the strengths of multiple existing approaches, accomplishingsignificant performance improvements over previous state of the art. Secondly, we introducean effective branching strategy on ReLU non-linearities. This branching strategy allows usto efficiently and successfully deal with high input dimensional problems with convolutionalnetwork architecture, on which previous methods fail frequently. Finally, we proposecomprehensive test data sets and benchmarks which includes a collection of previouslyreleased testcases. We use the data sets to conduct a thorough experimental comparison ofexisting and new algorithms and to provide an inclusive analysis of the factors impactingthe hardness of verification problems.

Machine learning is used extensively in recommender systems deployed in products. The decisions made by these systems can influence user beliefs and preferences which in turn affect the feedback the learning system receives - thus creating a feedback loop. This phenomenon can give rise to the so-called "echo chambers" or "filter bubbles" that have user and societal implications. In this paper, we provide a novel theoretical analysis that examines both the role of user dynamics and the behavior of recommender systems, disentangling the echo chamber from the filter bubble effect. In addition, we offer practical solutions to slow down system degeneracy. Our study contributes toward understanding and developing solutions to commonly cited issues in the complex temporal scenario, an area that is still largely unexplored.

Neural architectures able to generate variable-length outputs are extremely effective for applications like Machine Translation and Image Captioning. In this paper, we study the vulnerability of these models to attacks aimed at changing the output-size that can have undesirable consequences including increased computation and inducing faults in downstream modules that expect outputs of a certain length. We show the existence and construction of such attacks with two key contributions. First, to overcome the difficulties of discrete search space and the non-differentiable adversarial objective function, we develop an easy-to-compute differentiable proxy objective that can be used with gradient-based algorithms to find output-lengthening inputs. Second, we develop a verification approach to formally prove that the network cannot produce outputs greater than a certain length. Experimental results on Machine Translation and Image Captioning models show that our adversarial output-lengthening approach can produce outputs that are 50 times longer than the input, while our verification approach can, given a model and input domain, prove that the output length is below a certain size.

Po-Sen Huang, Robert Stanforth, Johannes Welbl, Chris Dyer, Dani Yogatama, Sven Gowal, Krishnamurthy Dvijotham, Pushmeet Kohli. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Recent work has shown that it is possible to train deep neural networks that are provably robust to norm-bounded adversarial perturbations. Most of these methods are based on minimizing an upper bound on the worst-case loss over all possible adversarial perturbations. While these techniques show promise, they often result in difficult optimization procedures that remain hard to scale to larger networks. Through a comprehensive analysis, we show how a simple bounding technique, interval bound propagation (IBP), can be exploited to train large provably robust neural networks that beat the state-of-the-art in verified accuracy. While the upper bound computed by IBP can be quite weak for general networks, we demonstrate that an appropriate loss and clever hyper-parameter schedule allow the network to adapt such that the IBP bound is tight. This results in a fast and stable learning algorithm that outperforms more sophisticated methods and achieves state-of-the-art results on MNIST, CIFAR-10 and SVHN. It also allows us to train the largest model to be verified beyond vacuous bounds on a downscaled version of IMAGENET.

Physical construction—the ability to compose objects, subject to physical dynamics, to serve some function—is fundamental to human intelligence. We introduce a suite of challenging physical construction tasks inspired by how children play with blocks, such as matching a target configuration, stacking blocks to connect objects together, and creating shelter-like structures over target objects. We examine how a range of deep reinforcement learning agents fare on these challenges, and introduce several new approaches which provide superior performance. Our results show that agents which use structured representations (e.g., objects and scene graphs) and structured policies (e.g., object-centric actions) outperform those which use less structured representations, and generalize better beyond their training when asked to reason about larger scenes. Model-based agents which use Monte-Carlo Tree Search also outperform strictly model-free agents in our most challenging construction problems. We conclude that approaches which combine structured representations and reasoning with powerful learning are a key path toward agents that possess rich intuitive physics, scene understanding, and planning.

We introduce Compositional Imitation Learning and Execution (CompILE): a framework for learning reusable, variable-length segments of hierarchically-structured behavior from demonstration data. CompILE uses a novel unsupervised, fully-differentiable sequence segmentation module to learn latent encodings of sequential data that can be re-composed and executed to perform new tasks. Once trained, our model generalizes to sequences of longer length and from environment instances not seen during training. We evaluate CompILE in a challenging 2D multi-task environment and a continuous control task, and show that it can find correct task boundaries and event encodings in an unsupervised manner. Latent codes and associated behavior policies discovered by CompILE can be used by a hierarchical agent, where the high-level policy selects actions in the latent code space, and the low-level, task-specific policies are simply the learned decoders. We found that our CompILE-based agent could learn given only sparse rewards, where agents without task-specific policies struggle.

This paper addresses the challenging problem of retrieval and matching of graph structured objects, and makes two key contributions. First, we demonstrate how Graph Neural Networks (GNN), which have emerged as an effective model for various supervised prediction problems defined on structured data, can be trained to produce embedding of graphs in vector spaces that enables efficient similarity reasoning. Second, we propose a novel Graph Matching Network model that, given a pair of graphs as input, computes a similarity score between them by jointly reasoning on the pair through a new cross-graph attention-based matching mechanism. We demonstrate the effectiveness of our models on different domains including the challenging problem of control-flow graph based function similarity search that plays an important role in the detection of vulnerabilities in software systems. The experimental analysis demonstrates that our models are not only able to exploit structure in the context of similarity learning but they can also outperform domain specific baseline systems that have been carefully hand-engineered for these problems.

This paper addressed the problem of formally verifying desirable properties of neural networks, i.e., obtaining provable guarantees that neural networks satisfy specifications relating their inputs and outputs (e.g., robustness to bounded norm adversarial perturbations). Most previous work on this topic was limited in its applicability by the size of the network, network architecture and the complexity of properties to be verified. In contrast, our framework applies to a general class of activation functions and specifications. We formulate verification as an optimization problem (seeking to find the largest violation of the specification) and solve a Lagrangian relaxation of the optimization problem to obtain an upper bound on the worst case violation of the specification being verified. Our approach is anytime, i.e., it can be stopped at any time and a valid bound on the maximum violation can be obtained. Finally, we highlight how this approach can be used to train models that are amenable to verification.

Recent work has uncovered the interesting (and somewhat surprising) finding that training models to be invariant to adversarial perturbations requires substantially larger datasets than those required for standard classification. This result is a key hurdle in the deployment of robust machine learning models in many real world applications where labeled data is expensive. Our main insight is that unlabeled data can be a competitive alternative to labeled data for training adversarially robust models. Theoretically, we show that in a simple statistical setting, the sample complexity for learning an adversarially robust model from unlabeled data matches the fully supervised case up to constant factors. On standard datasets like CIFAR- 10, a simple Unsupervised Adversarial Training (UAT) approach using unlabeled data improves robust accuracy by 21.7% over using 4K supervised examples alone, and captures over 95% of the improvement from the same number of labeled examples. Finally, we report an improvement of 4% over the previous state-of-the- art on CIFAR-10 against the strongest known attack by using additional unlabeled data from the uncurated 80 Million Tiny Images dataset. This demonstrates that our finding extends as well to the more realistic case where unlabeled data is also uncurated, therefore opening a new avenue for improving adversarial training.

This paper considers the problem of efficient exploration of unseen environments, a key challenge in AI. We propose a learning to explore' framework where we learn a policy from a distribution of environments. At test time, presented with an unseen environment from the same distribution, the policy aims to generalize the exploration strategy to visit the maximum number of unique states in a limited number of steps. We particularly focus on environments with graph-structured state-spaces that are encountered in many important real-world applications like software testing and map building. We formulate this task as a reinforcement learning problem where theexploration' agent is rewarded for transitioning to previously unseen environment states and employ a graph-structured memory to encode the agent's past trajectory. Experimental results demonstrate that our approach is extremely effective for exploration of spatial maps; and when applied on the challenging problems of coverage-guided software-testing of domain-specific programs and real-world mobile applications, it outperforms methods that have been hand-engineered by human experts.

Adversarial training is an effective methodology for training deep neural networks that are robust against adversarial, norm-bounded perturbations. However, the computational cost of adversarial training grows prohibitively as the size of the model and number of input dimensions increase. Further, training against less expensive and therefore weaker adversaries produces models that are robust against weak attacks but break down under attacks that are stronger. This is often attributed to the phenomenon of gradient obfuscation; such models have a highly non-linear loss surface in the vicinity of training examples, making it hard for gradient-based attacks to succeed even though adversarial examples still exist. In this work, we introduce a novel regularizer that encourages the loss to behave linearly in the vicinity of the training data, thereby penalizing gradient obfuscation while encouraging robustness. We show via extensive experiments on CIFAR-10 and ImageNet, that models trained with our regularizer avoid gradient obfuscation and can be trained significantly faster than adversarial training. Using this regularizer, we exceed current state of the art and achieve 47% adversarial accuracy for ImageNet with L-infinity norm adversarial perturbations of radius 4/255 under an untargeted, strong, white-box attack. Additionally, we match state of the art results for CIFAR-10 at 8/255.

Hand pose estimation, formulated as an inverse problem, is typically optimized by an energy function over pose parameters using a 'black box' image generation procedure, knowing little about either the relationships between the parameters or the form of the energy function. In this paper, we show significant improvement upon such black box optimization by exploiting high-level knowledge of the parameter structure and using a local surrogate energy function. Our new framework, hierarchical sampling optimization (HSO), consists of a sequence of discriminative predictors organized into a kinematic hierarchy. Each predictor is conditioned on its ancestors, and generates a set of samples over a subset of the pose parameters, with only one selected by the highly-efficient surrogate energy. The selected partial poses are concatenated to generate a full-pose hypothesis. Repeating the same process, several hypotheses are generated and the full energy function selects the best result. Under the same kinematic hierarchy, two methods based on decision forest and convolutional neural network are proposed to generate the samples and two optimization methods are studied when optimizing these samples. Experimental evaluations on three publicly available datasets show that our method is particularly impressive in low-compute scenarios where it significantly outperforms all other state-of-the-art methods.

Dense conditional random fields (CRFs) have become a popular framework for modeling several problems in computer vision such as stereo correspondence and multiclass semantic segmentation. By modeling long-range interactions, dense CRFs provide a labeling that captures finer detail than their sparse counterparts. Currently, the state-of-the-art algorithm performs mean-field inference using a filter-based method but fails to provide a strong theoretical guarantee on the quality of the solution. A question naturally arises as to whether it is possible to obtain a maximum a posteriori (MAP) estimate of a dense CRF using a principled method. Within this paper, we show that this is indeed possible. Specifically, we will show that, by using a filter-based method, continuous relaxations of the MAP problem can be optimized efficiently using state-of-the-art algorithms. Specifically, we will solve a quadratic programming relaxation using the Frank--Wolfe algorithm and a linear programming relaxation by developing a proximal minimization framework. By exploiting labeling consistency in the higher-order potentials and utilizing the filter-based method, we are able to formulate the above algorithms such that each iteration has a complexity linear in the number of classes and random variables. The presented algorithms can be applied to any labeling problem using a dense CRF with sparse higher-order potentials. In this paper, we use semantic segmentation as an example application as it demonstrates the ability of the algorithm to scale to dense CRFs with large dimensions. We perform experiments on the Pascal dataset to indicate that the presented algorithms are able to attain lower energies than the mean-field inference method.

Bayesian optimization (BO) has become an effective approach for black-box function optimization problems when function evaluations are expensive and the optimum can be achieved within a relatively small number of queries. However, many cases, such as the ones with high-dimensional inputs, may require a much larger number of observations for optimization. Despite an abundance of observations thanks to parallel experiments, current BO techniques have been limited to merely a few thousand observations. In this paper, we propose ensemble Bayesian optimization (EBO) to address three current challenges in BO simultaneously: (1) large-scale observations; (2) high dimensional input spaces; and (3) selections of batch queries that balance quality and diversity. The key idea of EBO is to operate on an ensemble of additive Gaussian process models, each of which possesses a randomized strategy to divide and conquer. We show unprecedented, previously impossible results of scaling up BO to tens of thousands of observations within minutes of computation.

This paper investigates recently proposed approaches for defending against adversarial examples and evaluating adversarial robustness. We motivate adversarial risk as an objective for achieving models robust to worst-case inputs. We then frame commonly used attacks and evaluation metrics as defining a tractable surrogate objective to the true adversarial risk. This suggests that models may optimize this surrogate rather than the true adversarial risk. We formalize this notion as obscurity to an adversary, and develop tools and heuristics for identifying obscured models and designing transparent models. We demonstrate that this is a significant problem in practice by repurposing gradient-free optimization techniques into adversarial attacks, which we use to decrease the accuracy of several recently proposed defenses to near zero. Our hope is that our formulations and results will help researchers to develop more powerful defenses.

We present a reinforcement learning framework, called Programmatically Interpretable Reinforcement Learning (PIRL), that is designed to generate interpretable and verifiable agent policies. Unlike the popular Deep Reinforcement Learning (DRL) paradigm, which represents policies by neural networks, PIRL represents policies using a high-level, domain-specific programming language. Such programmatic policies have the benefits of being more easily interpreted than neural networks, and being amenable to verification by symbolic methods. We propose a new method, called Neurally Directed Program Search (NDPS), for solving the challenging nonsmooth optimization problem of finding a programmatic policy with maximal reward. NDPS works by first learning a neural policy network using DRL, and then performing a local search over programmatic policies that seeks to minimize a distance from this neural “oracle”. We evaluate NDPS on the task of learning to drive a simulated car in the TORCS car-racing environment. We demonstrate that NDPS is able to discover human-readable policies that pass some significant performance bars. We also show that PIRL policies can have smoother trajectories, and can be more easily transferred to environments not encountered during training, than corresponding policies discovered by DRL.

Automatic correction of programs is a challenging problem with numerous real world applications in security, verification, and education. One application that is becoming increasingly important is the correction of student submissions in online courses for providing feedback. Most existing program repair techniques analyze Abstract Syntax Trees (ASTs) of programs, which are unfortunately unavailable for programs with syntax errors. In this paper, we propose a novel Neuro-symbolic approach that combines neural networks with constraint-based reasoning. Specifically, our method first uses a Recurrent Neural Network (RNN) to perform syntax repairs for the buggy programs; subsequently, the resulting syntactically-fixed programs are repaired using constraint-based techniques to ensure functional correctness. The RNNs are trained using a corpus of syntactically correct submissions for a given programming assignment, and are then queried to fix syntax errors in an incorrect programming submission by replacing or inserting the predicted tokens at the error location. We evaluate our technique on a dataset comprising of over 14,500 student submissions with syntax errors. Our method is able to repair syntax errors in 60% (8689) of submissions, and finds functionally correct repairs for 23.8% (3455) submissions.

The success of Deep Learning and its potential use in many safety-critical applications has motivated research on formal verification of Neural Network (NN) models. Despite the reputation of learned NN models to behave as black boxes and the theoretical hardness of proving their properties, researchers have been successful in verifying some classes of models by exploiting their piecewise linear structure and taking insights from formal methods such as Satisifiability Modulo Theory. These methods are however still far from scaling to realistic neural networks. To facilitate progress on this crucial area, we make two key contributions. First, we present a unified framework that encompasses previous methods. This analysis results in the identification of new methods that combine the strengths of multiple existing approaches, accomplishing a speedup of two orders of magnitude compared to the previous state of the art. Second, we propose a new data set of benchmarks which includes a collection of previously released testcases. We use the benchmark to provide the first experimental comparison of existing algorithms and identify the factors impacting the hardness of verification problems.

We marry two powerful ideas: deep representation learning for visual recognition and language understanding, and symbolic program execution for reasoning. Our neural-symbolic visual question answering (NS-VQA) system first recovers a structural scene representation from the image and a program trace from the question. It then executes the program on the scene representation to obtain an answer. Incorporating symbolic structure as prior knowledge offers three unique advantages. First, executing programs on a symbolic space is more robust to long program traces; our model can solve complex reasoning tasks better, achieving an accuracy of 99.8% on the CLEVR dataset. Second, the model is more data- and memory-efficient: it performs well after learning on a small number of training data; it can also encode an image into a compact representation, requiring less storage than existing methods for offline question answering. Third, symbolic program execution offers full transparency to the reasoning process; we are thus able to interpret and diagnose each execution step.

We study the problem of holistic scene understanding. We would like to obtain a compact, expressive, and interpretable representation of scenes that encodes information such as the number of objects and their categories, poses, positions, etc. Such a representation would allow us to reason about and even reconstruct or manipulate elements of the scene. Previous works have used encoder-decoder based neural architectures to learn image representations, however, representations obtained in this way are typically uninterpretable, or only explain a single object in the scene. In this work, we propose a new approach to learn an interpretable distributed representation of scenes. Our approach employs a deterministic rendering function as the decoder, mapping a naturally structured and disentangled scene description, which we named scene XML, to an image. By doing so, the encoder is forced to perform the inverse of the rendering operation (a.k.a. de-rendering) to transform an input image to the structured scene XML that the decoder used to produce the image. We use a object proposal based encoder that is trained by minimizing both the supervised prediction and the unsupervised reconstruction errors. Experiments demonstrate that our approach works well on scene de-rendering with two different graphics engines, and our learned representation can be easily adapted for a wide range of applications like image editing, inpainting, visual analogy-making, and image captioning.

This paper addresses the problem of converting a rasterized floorplan image into a vector-graphics representation. Unlike existing approaches that rely on a sequence of lowlevel image processing heuristics, we adopt a learning-based approach. A neural architecture first transforms a rasterized image to a set of junctions that represent low-level geometric and semantic information (e.g., wall corners or door end-points). Integer programming is then formulated to aggregate junctions into a set of simple primitives (e.g., wall lines, door lines, or icon boxes) to produce a vectorized floorplan, while ensuring a topologically and geometrically consistent result. Our algorithm significantly outperforms existing methods and achieves around 90% precision and recall, getting to the range of production-ready performance. The vector representation allows 3D model popup for better indoor scene visualization, direct model manipulation for architectural remodeling, and further computational applications such as data analysis. Our system is efficient: we have converted hundred thousand production-level floorplan images into the vector representation and generated 3D popup models.

We present a novel method to realistically puppeteer and animate a face from a single RGB image using a source video sequence. We begin by fitting a multilinear PCA model to obtain the 3D geometry and a single texture of the target face. In order for the animation to be realistic, however, we need dynamic per-frame textures that capture subtle wrinkles and deformations corresponding to the animated facial expressions. This problem is highly underconstrained, as dynamic textures cannot be obtained directly from a single image. Furthermore, if the target face has a closed mouth, it is not possible to obtain actual images of the mouth interior. To address this issue, we train a Deep Generative Network that can infer realistic per-frame texture deformations, including the mouth interior, of the target identity using the per-frame source textures and the single target texture. By retargeting the PCA expression geometry from the source, as well as using the newly inferred texture, we can both animate the face and perform video face replacement on the source video using the target appearance.

3D context has been shown to be extremely important for scene understanding, yet very little research has been done on integrating context information with deep neural network architectures. This paper presents an approach to embed 3D context into the topology of a neural network trained to perform holistic scene understanding. Given a depth image depicting a 3D scene, our network aligns the observed scene with a predefined 3D scene template, and then reasons about the existence and location of each object within the scene template. In doing so, our model recognizes multiple objects in a single forward pass of a 3D convolutional neural network, capturing both global scene and local object information simultaneously. To create training data for this 3D network, we generate partially synthetic depth images which are rendered by replacing real objects with a repository of CAD models of the same object category1. Extensive experiments demonstrate the effectiveness of our algorithm compared to the state of the art.

Combining abstract, symbolic reasoning with continuous neural reasoning is a grand challenge of representation learning. As a step in this direction, we propose a new architecture, called neural equivalence network, for the problem of learning continuous semantic representations of algebraic and logical expressions. These networks are trained to represent semantic equivalence, even of expressions that are syntactically very different. The challenge is that semantic representations must be computed in a syntax-directed manner, because semantics is compositional, but at the same time, small changes in syntax can lead to very large changes in semantics, which can be difficult for continuous neural architectures. We perform an exhaustive evaluation on the task of checking equivalence on a highly diverse class of symbolic algebraic and boolean expression types, showing that our model significantly outperforms existing architectures.

The problem of automatically generating a computer program from some specification has been studied since the early days of AI. Recently, two competing approaches for `automatic program learning’ have received significant attention: (1) `neural program synthesis’, where a neural network is conditioned on input/output (I/O) examples and learns to generate a program, and (2) `neural program induction’, where a neural network generates new outputs directly using a latent program representation. Here, for the first time, we directly compare both approaches on a large-scale, real-world learning task and we additionally contrast to rule-based program synthesis, which uses hand-crafted semantics to guide the program generation. Our neural models use a modified attention RNN to allow encoding of variable-sized sets of I/O pairs, which achieve 92\% accuracy on a real-world test set, compared to the 34\% accuracy of the previous best neural synthesis approach. The synthesis model also outperforms a comparable induction model on this task, but we more importantly demonstrate that the strength of each approach is highly dependent on the evaluation metric and end-user application. Finally, we show that we can train our neural models to remain very robust to the type of noise expected in real-world data (e.g., typos), while a highly-engineered rule-based system fails entirely.

Many real-world problems, such as network packet routing and urban traffic control, are naturally modeled as multi-agent reinforcement learning (RL) problems. However, existing multi-agent RL methods typically scale poorly in the problem size. Therefore, a key challenge is to translate the success of deep learning on single-agent RL to the multi-agent setting. A major stumbling block is that independent Q-learning, the most popular multi-agent RL method, introduces nonstationarity that makes it incompatible with the experience replay memory on which deep Q-learning relies. This paper proposes two methods that address this problem: 1) using a multi-agent variant of importance sampling to naturally decay obsolete data and 2) conditioning each agent’s value function on a fingerprint that disambiguates the age of the data sampled from the replay memory. Results on a challenging decentralised variant of StarCraft unit micromanagement confirm that these methods enable the successful combination of experience replay with multi-agent RL.

As a step towards developing zero-shot task generalization capabilities in reinforcement learning (RL), we introduce a new RL problem where the agent should learn to execute sequences of instructions after learning useful skills that solve subtasks. In this problem, we consider two types of generalizations: to previously unseen instructions and to longer sequences of instructions. For generalization over unseen instructions, we propose a new objective which encourages learning correspondences between similar subtasks by making analogies. For generalization over sequential instructions, we present a hierarchical architecture where a meta controller learns to use the acquired skills for executing the instructions. To deal with delayed reward, we propose a new neural architecture in the meta controller that learns when to update the subtask, which makes learning more efficient. Experimental results on a stochastic 3D domain show that the proposed ideas are crucial for generalization to longer instructions as well as unseen instructions.

Optimization of high-dimensional black-box functions is an extremely challenging problem. While Bayesian optimization has emerged as a popular approach for optimizing black-box functions, its applicability has been limited to low-dimensional problems due to its computational and statistical challenges arising from high-dimensional settings. In this paper, we propose to tackle these challenges by (1) assuming a latent additive structure in the function and inferring it properly for more efficient and effective BO, and (2) performing multiple evaluations in parallel to reduce the number of iterations required by the method. Our novel approach learns the latent structure with Gibbs sampling and constructs batched queries using determinantal point processes. Experimental validations on both synthetic and real-world functions demonstrate that the proposed method outperforms the existing state-of-the-art approaches.

We introduce a paradigm for understanding physical scenes without human annotations. At the core of our system is a physical world representation that is first recovered by a perception module and then utilized by physics and graphics engines. During training, the perception module and the generative models learn by visual de-animation --- interpreting and reconstructing the visual information stream. During testing, the system first recovers the physical world state, and then uses the generative models for reasoning and future prediction. Even more so than forward simulation, inverting a physics or graphics engine is a computationally hard problem; we overcome this challenge by using a convolutional inversion network. Our system quickly recognizes the physical world state from appearance and motion cues, and has the flexibility to incorporate both differentiable and non-differentiable physics and graphics engines. We evaluate our system on both synthetic and real datasets involving multiple physical scenes, and demonstrate that our system performs well on both physical state estimation and reasoning problems. We further show that the knowledge learned on the synthetic dataset generalizes to constrained real images.

Most recently proposed methods for Neural Program induction work under the assumption of having a large set of input/output (I/O) examples for learning any given input-output mapping. This paper aims to address the problem of data and computation efficiency of program induction by leveraging information from related tasks. Specifically, we propose two novel approaches for cross-task knowledge transfer to improve program induction in limited-data scenarios. In our first proposal, portfolio adaptation, a set of induction models is pretrained on a set of related tasks, and the best model is adapted towards the new task using transfer learning. In our second approach, meta program induction, a $k$-shot learning approach is used to make a model generalize to new tasks without additional training. To test the efficacy of our methods, we constructed a new benchmark of programs written in the Karel programming language. Using an extensive experimental evaluation on the Karel benchmark, we demonstrate that our proposals dramatically outperform the baseline induction method that does not use knowledge transfer. We also analyze the relative performance of the two approaches and study conditions in which they perform best. In particular, meta induction outperforms all existing approaches under extreme data sparsity (when a very small number of examples are available), i.e., fewer than ten. As the number of available I/O examples increase (i.e. a thousand or more), portfolio adapted program induction becomes the best approach. For intermediate data sizes, we demonstrate that the combined method of adapted meta program induction has the strongest performance.

Variational autoencoders (VAEs) learn representations of data by jointly training a probabilistic encoder and decoder network. Typically these models encode all features of the data into a single variable. Here we are interested in learning disentangled representations that encode distinct aspects of the data into separate variables. We propose to learn such representations using model architectures that generalise from standard VAEs, employing a general graphical model structure in the encoder and decoder. This allows us to train partially-specified models that make relatively strong assumptions about a subset of interpretable variables and rely on the flexibility of neural networks to learn representations for the remaining variables. We further define a general objective for semi-supervised learning in this model class, which can be approximated using an importance sampling procedure. We evaluate our framework's ability to learn disentangled representations, both by qualitative exploration of its generative capacity, and quantitative evaluation of its discriminative ability on a variety of models and datasets.

In this paper, we present a novel, general, and efficient architecture for addressing computer vision problems that are approached from an 'Analysis by Synthesis' standpoint. Analysis by synthesis involves the minimization of reconstruction error, which is typically a non-convex function of the latent target variables. State-of-the-art methods adopt a hybrid scheme where discriminatively trained predictors like Random Forests or Convolutional Neural Networks are used to initialize local search algorithms. While these hybrid methods have been shown to produce promising results, they often get stuck in local optima. Our method goes beyond the conventional hybrid architecture by not only proposing multiple accurate initial solutions but by also defining a navigational structure over the solution space that can be used for extremely efficient gradient-free local search. We demonstrate the efficacy and generalizability of our approach on tasks as diverse as Hand Pose Estimation, RGB Camera Relocalization, and Image Retrieval.

This paper addresses the challenging problem of perceiving the hidden or occluded geometry of the scene depicted in any given RGBD image. Unlike other image labeling problems such as image segmentation where each pixel needs to be assigned a single label, layered decomposition requires us to assign multiple labels to pixels. We propose a novel "Occlusion-CRF" model that allows for the integration of sophisticated priors to regularize the solution space and enables the automatic inference of the layer decomposition. We use a generalization of the Fusion Move algorithm to perform Maximum a Posterior (MAP) inference on the model that can handle the large label sets needed to represent multiple surface assignments to each pixel. We have evaluated the proposed model and the inference algorithm on many RGBD images of cluttered indoor scenes. Our experiments show that not only is our model able to explain occlusions but it also enables automatic inpainting of occluded/ invisible surfaces.

This paper proposes a novel extremely efficient, fully-parallelizable, task-specific algorithm for the computation of global point-wise correspondences in images and videos. Our algorithm, the Global Patch Collider, is based on detecting unique collisions between image points using a collection of learned tree structures that act as conditional hash functions. In contrast to conventional approaches that rely on pairwise distance computation, our algorithm isolates distinctive pixel pairs that hit the same leaf during traversal through multiple learned tree structures. The split functions stored at the intermediate nodes of the trees are trained to ensure that only visually similar patches or their geometric or photometric transformed versions fall into the same leaf node. The matching process involves passing all pixel positions in the images under analysis through the tree structures. We then compute matches by isolating points that uniquely collide with each other ie. fell in the same empty leaf in multiple trees. Our algorithm is linear in the number of pixels but can be made constant time on a parallel computation architecture as the tree traversal for individual image points is decoupled. We demonstrate the efficacy of our method by using it to perform optical flow matching and stereo matching on some challenging benchmarks. Experimental results show that not only is our method extremely computationally efficient, but it is also able to match or outperform state of the art methods that are much more complex.

Ting-Hao Kenneth Huang, Francis Ferraro, Nasrin Mostafazadeh, Ishan Misra, Aishwarya Agrawal, Jacob Devlin, Ross Girshick, Xiaodong He, Pushmeet Kohli, Dhruv Batra, C. Lawrence Zitnick, Devi Parikh, Lucy Vanderwende, Michel Galley, Margaret Mitchell. Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2016.

Nasrin Mostafazadeh, Nathanael Chambers, Xiaodong He, Devi Parikh, Dhruv Batra, Lucy Vanderwende, Pushmeet Kohli, James Allen. Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2016.

This paper proposes an adaptive neural-compilation framework to address the problem of learning efficient program. Traditional code optimisation strategies used in compilers are based on applying pre-specified set of transformations that make the code faster to execute without changing its semantics. In contrast, our work involves adapting programs to make them more efficient while considering correctness only on a target input distribution. Our approach is inspired by the recent works on differentiable representations of programs. We show that it is possible to compile programs written in a low-level language to a differentiable representation. We also show how programs in this representation can be optimised to make them efficient on a target distribution of inputs. Experimental results demonstrate that our approach enables learning specifically-tuned algorithms for given data distributions with a high success rate.

We propose a novel approach to reduce the computational cost of evaluation of convolutional neural networks, a factor that has hindered their deployment in low-power devices such as mobile phones. Inspired by the loop perforation technique from source code optimization, we speed up the bottleneck convolutional layers by skipping their evaluation in some of the spatial positions. We propose and analyze several strategies of choosing these positions. We demonstrate that perforation can accelerate modern convolutional networks such as AlexNet and VGG-16 by a factor of 2x - 4x. Additionally, we show that perforation is complementary to the recently proposed acceleration method of Zhang et al.

Gaussian Process bandit optimization has emerged as a powerful tool for optimizing noisy black box functions. One example in machine learning is hyper-parameter optimization where each evaluation of the target function may require training a model which may involve days or even weeks of computation. Most methods for this so-called “Bayesian optimization” only allow sequential exploration of the parameter space. However, it is often desirable to propose batches or sets of parameter values to explore simultaneously, especially when there are large parallel processing facilities at our disposal. Batch methods require modeling the interaction between the different evaluations in the batch, which can be expensive in complex scenarios. In this paper, we propose a new approach for parallelizing Bayesian optimization by modeling the diversity of a batch via Determinantal point processes (DPPs) whose kernels are learned automatically. This allows us to generalize a previous result as well as prove better regret bounds based on DPP sampling. Our experiments on a variety of synthetic and real-world robotics and hyper-parameter optimization tasks indicate that our DPP-based methods, especially those based on DPP sampling, outperform state-of-the-art methods.

We present an end-to-end system for augmented and virtual reality telepresence, called Holoportation. Our system demonstrates high-quality, real-time 3D reconstructions of an entire space, including people, furniture and objects, using a set of new depth cameras. These 3D models can also be transmitted in real-time to remote users. This allows users wearing virtual or augmented reality displays to see, hear and interact with remote participants in 3D, almost as if they were present in the same physical space. From an audio-visual perspective, communicating and interacting with remote users edges closer to face-to-face communication. This paper describes the Holoportation technical system in full, its key interactive capabilities, the application scenarios it enables, and an initial qualitative study of using this new communication medium.

Many aspects of the design of efficient crowdsourcing processes, such as defining worker’s bonuses, fair prices and time limits of the tasks, involve knowledge of the likely duration of the task at hand. In this work we introduce a new time–sensitive Bayesian aggregation method that simultaneously estimates a task’s duration and obtains reliable aggregations of crowdsourced judgments. Our method, called BCCTime, uses latent variables to represent the uncertainty about the workers’ completion time, the tasks’ duration and the workers’ accuracy. To relate the quality of a judgment to the time a worker spends on a task, our model assumes that each task is completed within a latent time window within which all workers with a propensity to genuinely attempt the labelling task (i.e., no spammers) are expected to submit their judgments. In contrast, workers with a lower propensity to valid labelling, such as spammers, bots or lazy labellers, are assumed to perform tasks considerably faster or slower than the time required by normal workers. Specifically, we use efficient message-passing Bayesian inference to learn approximate posterior probabilities of (i) the confusion matrix of each worker, (ii) the propensity to valid labelling of each worker, (iii) the unbiased duration of each task and (iv) the true label of each task. Using two real- world public datasets for entity linking tasks, we show that BCCTime produces up to 11% more accurate classifications and up to 100% more informative estimates of a task’s duration compared to state–of–the–art methods.

We contribute a new pipeline for live multi-view performance capture, generating temporally coherent high-quality reconstructions in real-time. Our algorithm supports both incremental reconstruction, improving the surface estimation over time, as well as parameterizing the nonrigid scene motion. Our approach is highly robust to both large frame-to-frame motion and topology changes, allowing us to reconstruct extremely challenging scenes. We demonstrate advantages over related real-time techniques that either deform an online generated template or continually fuse depth data nonrigidly into a single reference model. Finally, we show geometric reconstruction results on par with offline methods which require orders of magnitude more processing time and many more RGBD cameras.

Fully articulated hand tracking promises to enable fundamentally new interactions with virtual and augmented worlds, but the limited accuracy and efficiency of current systems has prevented widespread adoption. Today's dominant paradigm uses machine learning for initialization and recovery followed by iterative model-fitting optimization to achieve a detailed pose fit. We follow this paradigm, but make several changes to the model-fitting, namely using: (1) a more discriminative objective function; (2) a smooth-surface model that provides gradients for non-linear optimization; and (3) joint optimization over both the model pose and the correspondences between observed data points and the model surface. While each of these changes may actually increase the cost per fitting iteration, we find a compensating decrease in the number of iterations. Further, the wide basin of convergence means that fewer starting points are needed for successful model fitting. Our system runs in real-time on CPU only, which frees up the commonly over-burdened GPU for experience designers. The hand tracker is efficient enough to run on low-power devices such as tablets. We can track up to several meters from the camera to provide a large working volume for interaction, even using the noisy data from current-generation depth cameras. Quantitative assessments on standard datasets show that the new approach exceeds the state of the art in accuracy. Qualitative results take the form of live recordings of a range of interactive experiences enabled by this new approach.

Generative models provide a powerful framework for probabilistic reasoning. However, in many domains their use has been hampered by the practical difficulties of inference. This is particularly the case in computer vision, where models of the imaging process tend to be large, loopy and layered. For this reason bottom-up conditional models have traditionally dominated in such domains. We find that widely-used, general-purpose message passing inference algorithms such as Expectation Propagation (EP) and Variational Message Passing (VMP) fail on the simplest of vision models. With these models in mind, we introduce a modification to message passing that learns to exploit their layered structure by passing ’consensus’ messages that guide inference towards good solutions. Experiments on a variety of problems show that the proposed technique leads to significantly more accurate inference results, not only when compared to standard EP and VMP, but also when compared to competitive bottom-up conditional models.

We present a new real-time hand tracking system based on a single depth camera. The system can accurately reconstruct complex hand poses across a variety of subjects. It also allows for robust tracking, rapidly recovering from any temporary failures. Most uniquely, our tracker is highly flexible, dramatically improving upon previous approaches which have focused on front-facing close-range scenarios. This flexibility opens up new possibilities for human-computer interaction with examples including tracking at distances from tens of centimeters through to several meters (for controlling the TV at a distance), supporting tracking using a moving depth camera (for mobile scenarios), and arbitrary camera placements (for VR headsets). These features are achieved through a new pipeline that combines a multi-layered discriminative reinitialization strategy for per-frame pose estimation, followed by a generative model-fitting stage. We provide extensive technical details and a detailed qualitative and quantitative analysis.

Recent progress on probabilistic modeling and statistical learning, coupled with the availability of large training datasets, has led to remarkable progress in computer vision. Generative probabilistic models, or “analysis-by-synthesis” approaches, can capture rich scene structure but have been less widely applied than their discriminative counterparts, as they often require considerable problem-specific engineering in modeling and inference, and inference is typically seen as requiring slow, hypothesize-and-test Monte Carlo methods. Here we present Picture, a probabilistic programming language for scene understanding that allows researchers to express complex generative vision models, while automatically solving them using fast general-purpose inference machinery. Picture provides a stochastic scene language that can express generative models for arbitrary 2D/3D scenes, as well as a hierarchy of representation layers for comparing scene hypotheses with observed images by matching not simply pixels, but also more abstract features (e.g., contours, deep neural network activations). Inference can flexibly integrate advanced Monte Carlo strategies with fast bottom-up data-driven methods. Thus both representations and inference strategies can build directly on progress in discriminatively trained systems to make generative vision more robust and efficient. We use Picture to write programs for 3D face analysis, 3D human pose estimation, and 3D object reconstruction - each competitive with specially engineered baselines.

The increase in size of large image databases makes the problem of efficient retrieval extremely challenging. This is especially true in the case of high dimensional data where even operations like hashing become expensive because of costly projection operators. Unlike most hashing methods that sacrifice accuracy for speed, we propose a novel method that improves the speed of high dimensional image retrieval by several orders of magnitude without any significant drop in performance. To do this, we propose to learn computationally bounded sparse projections for the encoding step. To further increase the accuracy of the method, we add an orthogonality constraint on projections to reduce bit correlation. We then introduce an iterative scheme that jointly optimizes this objective, which helps us obtain fast and efficient projections. We demonstrate this technique on large retrieval databases, specifically ImageNET, GIST1M and SUN-attribute for the task of nearest neighbor retrieval, and show that our method achieves a speed-up of up to a factor of 100 over state-of-the-art methods, while having on-par and in some cases even better accuracy.

This paper addresses the problem of learning long binary codes from high-dimensional data. We observe that two key challenges arise while learning and using long binary codes: (1) lack of an effective regularizer for the learned high-dimensional mapping and (2) high computational cost for computing long codes. In this paper, we overcome both these problems by introducing a sparsity encouraging regularizer that reduces the effective number of parameters involved in the learned projection operator. This regularizer not only reduces overfitting but, due to the sparse nature of the projection matrix, also leads to a dramatic reduction in the computational cost. To evaluate the effectiveness of our method, we analyze its performance on the problems of nearest neighbour search, image retrieval and image classification. Experiments on a number of challenging datasets show that our method leads to better accuracy than dense projections (ITQ [11] and LSH [16]) with the same code lengths, and meanwhile is over an order of magnitude faster. Furthermore, our method is also more accurate and faster than other recently proposed methods for speeding up high-dimensional binary encoding.

Crowdsourcing and human computation are being employed in sophisticated projects that require the solution of a heterogeneous set of tasks. We explore the challenge of composing or hiring an effective team from an available pool of applicants for performing tasks required for such projects on an ongoing basis. How can one optimally spend budget to learn the expertise of workers as part of recruiting a team? How can one exploit the similarities among tasks as well as underlying social ties or commonalities among the workers for faster learning? We tackle these decision-theoretic challenges by casting them as an instance of online learning for best action selection with side-observations. We present algorithms with PAC bounds on the required budget to hire a near-optimal team with high confidence. We evaluate our methodology on simulated problem instances using crowdsourcing data collected from the Upwork platform.

We address the problem of hand pose estimation, formulated as an inverse problem. Typical approaches optimize an energy function over pose parameters using a 'black box' image generation procedure. This procedure knows little about either the relationships between the parameters or the form of the energy function. In this paper, we show that we can significantly improving upon black box optimization by exploiting high-level knowledge of the structure of the parameters and using a local surrogate energy function. Our new framework, hierarchical sampling optimization, consists of a sequence of predictors organized into a kinematic hierarchy. Each predictor is conditioned on its ancestors, and generates a set of samples over a subset of the pose parameters. The highly-efficient surrogate energy is used to select among samples. Having evaluated the full hierarchy, the partial pose samples are concatenated to generate a full-pose hypothesis. Several hypotheses are generated using the same procedure, and finally the original full energy function selects the best result. Experimental evaluation on three publically available datasets show that our method is particularly impressive in low-compute scenarios where it significantly outperforms all other state-of-the-art methods.

This paper presents the Deep Convolution Inverse Graphics Network (DC-IGN), a model that aims to learn an interpretable representation of images, disentangled with respect to three-dimensional scene structure and viewing transformations such as depth rotations and lighting variations. The DC-IGN model is composed of multiple layers of convolution and de-convolution operators and is trained using the Stochastic Gradient Variational Bayes (SGVB) algorithm. We propose a training procedure to encourage neurons in the graphics code layer to represent a specific transformation (e.g. pose or light). Given a single input image, our model can generate new images of the same object with variations in pose and lighting. We present qualitative and quantitative tests of the model's efficacy at learning a 3D rendering engine for varied object classes including faces and chairs.

Decision trees and randomized forests are widely used in computer vision and machine learning. Standard algorithms for decision tree induction optimize the split functions one node at a time according to some splitting criteria. This greedy procedure often leads to suboptimal trees. In this paper, we present an algorithm for optimizing the split functions at all levels of the tree jointly with the leaf parameters, based on a global objective. We show that the problem of finding optimal linear-combination (oblique) splits for decision trees is related to structured prediction with latent variables, and we formulate a convex-concave upper bound on the tree's empirical loss. Computing the gradient of the proposed surrogate objective with respect to each training exemplar is O(d^2), where d is the tree depth, and thus training deep trees is feasible. The use of stochastic gradient descent for optimization enables effective training with large datasets. Experiments on several classification benchmarks demonstrate that the resulting non-greedy decision trees outperform greedy decision tree baselines.

Truncated signed distance function (TSDF) based volumetric surface reconstructions of static environments can be readily acquired using recent RGB-D camera based mapping systems. If objects in the environment move then a previously obtained TSDF reconstruction is no longer current. Handling this problem requires segmenting moving objects from the reconstruction. To this end, we present a novel solution to the motion segmentation of TSDF volumes. The segmentation problem is cast as CRF-based MAP inference in the voxel space. We propose: a novel data term by solving sparse multi-body motion segmentation and computing likelihoods for each motion label in the RGB-D image space, and, a novel pairwise term based on gradients of the TSDF volume. Experimental evaluation shows that the proposed approach achieves successful segmentations on reconstructions acquired with Kinect Fusion. Unlike the existing solutions which only work if the objects move completely from their initially occupied spaces, the proposed method permits segmentation of objects when they start to move.

Social media has led to the democratisation of opinion sharing. A wealth of information about public opinions, current events, and authors' insights into specific topics can be gained by understanding the text written by users. However, there is a wide variation in the language used by different authors in different contexts on the web. This diversity in language makes interpretation an extremely challenging task. Crowdsourcing presents an opportunity to interpret the sentiment, or topic, of free-text. However, the subjectivity and bias of human interpreters raise challenges in inferring the semantics expressed by the text. To overcome this problem, we present a novel Bayesian approach to language understanding that relies on aggregated crowdsourced judgements. Our model encodes the relationships between labels and text features in documents, such as tweets, web articles, and blog posts, accounting for the varying reliability of human labellers. It allows inference of annotations that scales to arbitrarily large pools of documents. Our evaluation using two challenging crowdsourcing datasets shows that by efficiently exploiting language models learnt from aggregated crowdsourced labels, we can provide up to 25% improved classifications when only a small portion, less than 4% of documents has been labelled. Compared to the six state-of-the-art methods, we reduce by up to 67% the number of crowd responses required to achieve comparable accuracy. Our method was a joint winner of the CrowdFlower - CrowdScale 2013 Shared Task challenge at the conference on Human Computation and Crowdsourcing (HCOMP 2013).

Abstract Detecting changes in scenes is important in many scene understanding tasks. In this paper, we pursue this goal simply from a pair of image recordings. Specifically, our goal is to infer what the objects are, how they are structured, and how they moved between the images. The problem is challenging as large changes make point‐level correspondence establishment difficult, which in turn breaks the assumptions of standard Structure‐from‐Motion (SfM). We propose a novel algorithm for dynamic SfM wherein we first generate a pool of potential corresponding points by hypothesizing over possible movements, and then use a continuous optimization formulation to obtain a low complexity solution that best explains the scene recordings, i.e., the input image pairs. We test the algorithm on a variety of examples to recover the multiple object structures and their changes.

We present a new interactive and online approach to 3D scene understanding. Our system, SemanticPaint , allows users to simultaneously scan their environment whilst interactively segmenting the scene simply by reaching out and touching any desired object or surface. Our system continuously learns from these segmentations, and labels new unseen parts of the environment. Unlike offline systems where capture, labeling, and batch learning often take hours or even days to perform, our approach is fully online. This provides users with continuous live feedback of the recognition during capture, allowing to immediately correct errors in the segmentation and/or learning—a feature that has so far been unavailable to batch and offline methods. This leads to models that are tailored or personalized specifically to the user's environments and object classes of interest, opening up the potential for new applications in augmented reality, interior design, and human/robot navigation. It also provides the ability to capture substantial labeled 3D datasets for training large-scale visual recognition systems.

We present the first pipeline for real-time volumetric surface reconstruction and dense 6DoF camera tracking running purely on standard, off-the-shelf mobile phones. Using only the embedded RGB camera, our system allows users to scan objects of varying shape, size, and appearance in seconds, with real-time feedback during the capture process. Unlike existing state of the art methods, which produce only point-based 3D models on the phone, or require cloud-based processing, our hybrid GPU/CPU pipeline is unique in that it creates a connected 3D surface model directly on the device at 25Hz. In each frame, we perform dense 6DoF tracking, which continuously registers the RGB input to the incrementally built 3D model, minimizing a noise aware photoconsistency error metric. This is followed by efficient key-frame selection, and dense per-frame stereo matching. These depth maps are fused volumetrically using a method akin to KinectFusion, producing compelling surface models. For each frame, the implicit surface is extracted for live user feedback and pose estimation. We demonstrate scans of a variety of objects, and compare to a Kinect-based baseline, showing on average ∼ 1.5cm error. We qualitatively compare to a state of the art point-based mobile phone method, demonstrating an order of magnitude faster scanning times, and fully connected surface models.

This paper contributes a real time method for recovering facial shape and expression from a single depth image. The method also estimates an accurate and dense correspondence field between the input depth image and a generic face model. Both outputs are a result of minimizing the error in reconstructing the depth image, achieved by applying a set of identity and expression blend shapes to the model. Traditionally, such a generative approach has shown to be computationally expensive and non-robust because of the non-linear nature of the reconstruction error. To overcome this problem, we use a discriminatively trained prediction pipeline that employs random forests to generate an initial dense but noisy correspondence field. Our method then exploits a fast ICP-like approximation to update these correspondences, allowing us to quickly obtain a robust initial fit of our model. The model parameters are then fine tuned to minimize the true reconstruction error using a stochastic optimization technique. The correspondence field resulting from our hybrid generative-discriminative pipeline is accurate and useful for a variety of applications such as mesh deformation and retexturing. Our method works in real-time on a single depth image i.e. Without temporal tracking, is free from per-user calibration, and works in low-light conditions.

This paper proposes a novel method for efficiently generating multiple diverse predictions for structured prediction problems. Existing methods like SDPPs or DivMBest work by making a series of predictions where each prediction is made after considering the predictions that came before it. Such approaches are inherently sequential and computationally expensive. In contrast, our method, Diverse Multiple Choice Learning, learns a set of models to make multiple independent, yet diverse, predictions at testtime. We achieve this by including a diversity encouraging term in the loss function used for training the models. This approach encourages diversity in the predictions while preserving computational efficiency at test-time. Experimental results on a number of challenging problems show that our method learns models that not only predict more diverse results than competing methods, but are also able to generalize better and produce results with high test accuracy.

We propose 'filter forests' (FF), an efficient new discriminative approach for predicting continuous variables given a signal and its context. FF can be used for general signal restoration tasks that can be tackled via convolutional filtering, where it attempts to learn the optimal filtering kernels to be applied to each data point. The model can learn both the size of the kernel and its values, conditioned on the observation and its spatial or temporal context. We show that FF compares favorably to both Markov random field based and recently proposed regression forest based approaches for labeling problems in terms of efficiency and accuracy. In particular, we demonstrate how FF can be used to learn optimal denoising filters for natural images as well as for other tasks such as depth image refinement, and 1D signal magnitude estimation. Numerous experiments and quantitative comparisons show that FFs achieve accuracy at par or superior to recent state of the art techniques, while being several orders of magnitude faster.

We address the problem of estimating the pose of a cam- era relative to a known 3D scene from a single RGB-D frame. We formulate this problem as inversion of the generative rendering procedure, i.e., we want to find the camera pose corresponding to a rendering of the 3D scene model that is most similar with the observed input. This is a non-convex optimization problem with many local optima. We propose a hybrid discriminative-generative learning architecture that consists of: (i) a set of M predictors which generate M camera pose hypotheses, and (ii) a 'selector' or 'aggregator' that infers the best pose from the multiple pose hypotheses based on a similarity function. We are interested in predictors that not only produce good hypotheses but also hypotheses that are different from each other. Thus, we propose and study methods for learning 'marginally relevant' predictors, and compare their performance when used with different selection procedures. We evaluate our method on a recently released 3D reconstruction dataset with challenging camera poses, and scene variability. Experiments show that our method learns to make multiple predictions that are marginally relevant and can effectively select an accurate prediction. Furthermore, our method outperforms the state-of-the-art discriminative approach for camera relocalization.

Human gestures, similar to speech and handwriting, are often unique to the individual. Training a generic classifier applicable to everyone can be very difficult and as such, it has become a standard to use personalized classifiers in speech and handwriting recognition. In this paper, we address the problem of personalization in the context of gesture recognition, and propose a novel and extremely efficient way of doing personalization. Unlike conventional personalization methods which learn a single classifier that later gets adapted, our approach learns a set (portfolio) of classifiers during training, one of which is selected for each test subject based on the personalization data. We formulate classifier personalization as a selection problem and propose several algorithms to compute the set of candidate classifiers. Our experiments show that such an approach is much more efficient than adapting the classifier parameters but can still achieve comparable or better results.

Creating a layout for an augmented reality (AR) application which embeds virtual objects in a physical environment is difficult as it must adapt to any physical space. We propose a rule-based framework for generating object layouts for AR applications. Under our framework, the developer of an AR application specifies a set of rules (constraints) which enforce self-consistency (rules regarding the inter-relationships of application components) and scene-consistency (application components are consistent with the physical environment they are placed in). When a user enters a new environment, we create, in real-time, a layout for the application, which is consistent with the defined constraints (as much as possible). We find the optimal configurations for each object by solving a constraint-satisfaction problem. Our stochastic move making algorithm is domain-aware, and allows us to efficiently converge to a solution for most rule-sets. In the paper we demonstrate several augmented reality applications that automatically adapt to different rooms and changing circumstances in each room.

Different interfaces allow a user to achieve the same end goal through different action sequences, e.g., command lines vs. drop down menus. Interface efficiency can be described in terms of a cost incurred, e.g., time taken, by the user in typical tasks. Realistic users arrive at evaluations of efficiency, hence making choices about which interface to use, over time, based on trial and error experience. Their choices are also determined by prior experience, which determines how much learning time is required. These factors have substantial effect on the adoption of new interfaces. In this paper, we aim at understanding how users adapt under interface change, how much time it takes them to learn to interact optimally with an interface, and how this learning could be expedited through intermediate interfaces. We present results from a series of experiments that make four main points: (a) different interfaces for accomplishing the same task can elicit significant variability in performance, (b) switching interfaces can result in adverse sharp shifts in performance, (c) subject to some variability, there are individual thresholds on tolerance to this kind of performance degradation with an interface, causing users to potentially abandon what may be a pretty good interface, and (d) our main result -- shaping user learning through the presentation of intermediate interfaces can mitigate the adverse shifts in performance while still enabling the eventual improved performance with the complex interface upon the user becoming suitably accustomed. In our experiments, human users use keyboard based interfaces to navigate a simulated ball through a maze. Our results are a first step towards interface adaptation algorithms that architect choice to accommodate personality traits of realistic users.

This paper addresses the problem of extracting accurate labels from crowdsourced datasets, a key challenge in crowdsourcing. Prior work has focused on modeling the reliability of individual workers, for instance, by way of confusion matrices, and using these latent traits to estimate the true labels more accurately. However, this strategy becomes ineffective when there are too few labels per worker to reliably estimate their quality. To mitigate this issue, we propose a novel community-based Bayesian label aggregation model, CommunityBCC, which assumes that crowd workers conform to a few different types, where each type represents a group of workers with similar confusion matrices. We assume that each worker belongs to a certain community, where the worker's confusion matrix is similar to (a perturbation of) the community's confusion matrix. Our model can then learn a set of key latent features: (i) the confusion matrix of each community, (ii) the community membership of each user, and (iii) the aggregated label of each item. We compare the performance of our model against established aggregation methods on a number of large-scale, real-world crowdsourcing datasets. Our experimental results show that our CommunityBCC model consistently outperforms state-of-the-art label aggregation methods, requiring, on average, 50% less data to pass the 90% accuracy mark.

Abstract In this paper, we propose a new method for reconstructing 3D models from a noisy and incomplete 3D scan and a coarse template model. The main idea is to maintain characteristic high‐level features of the template that remain unchanged for different variants of the same type of object. As invariants, we chose the partial symmetry structure of the template model under Euclidian transformations, i.e. we maintain the algebraic structure of all reflections, rotations and translations that map the object partially to itself. We propose an optimization scheme that maintains continuous and discrete symmetry properties of this kind while registering a template against scan data using a deformable iterative closest points (ICP) framework with thin‐plate‐spline regularization. We apply our new deformation approach to a large number of example data sets and demonstrate that symmetry‐guided template matching often yields much more plausible reconstructions than previous variants of ICP.

Abstract In this paper, we address the problem of structure‐aware shape deformation: We specifically consider deformations that preserve symmetries of the shape being edited. While this is an elegant approach for obtaining plausible shape variations from minimal assumptions, a straightforward optimization is numerically expensive and poorly conditioned. Our paper introduces an explicit construction of bases of linear spaces of shape deformations that exactly preserve symmetries for any user‐defined level of detail. This permits the construction of low‐dimensional spaces of low‐frequency deformations that preserve the symmetries. We obtain substantial speed‐ups over alternative approaches for symmetry‐preserving shape editing due to (i) the sub‐space approach, which permits low‐res editing, (ii) the removal of redundant, symmetric information, and (iii) the simplification of the numerical formulation due to hard‐coded symmetry preservation. We demonstrate the utility in practice by applying our framework to symmetry‐preserving co‐rotated iterative Laplace surface editing of models with complex symmetry structure, including partial and nested symmetry.

Individual differences in personality affect users’ online activities as much as they do in the offline world. This work, based on a sample of over a third of a million users, examines how users’ behaviour in the online environment, captured by their website choices and Facebook profile features, relates to their personality, as measured by the standard Five Factor Model personality questionnaire. Results show that there are psychologically meaningful links between users’ personalities, their website preferences and Facebook profile features. We show how website audiences differ in terms of their personality, present the relationships between personality and Facebook profile features, and show how an individual’s personality can be predicted from Facebook profile features. We conclude that predicting a user’s personality profile can be applied to personalize content, optimize search results, and improve online advertising.

In this paper, a hypergraph-based image segmentation framework is formulated in a supervised manner for many high-level computer vision tasks. To consider short- and long-range dependency among various regions of an image and also to incorporate wider selection of features, a higher-order correlation clustering (HO-CC) is incorporated in the framework. Correlation clustering (CC), which is a graph-partitioning algorithm, was recently shown to be effective in a number of applications such as natural language processing, document clustering, and image segmentation. It derives its partitioning result from a pairwise graph by optimizing a global objective function such that it simultaneously maximizes both intra-cluster similarity and inter-cluster dissimilarity. In the HO-CC, the pairwise graph which is used in the CC is generalized to a hypergraph which can alleviate local boundary ambiguities that can occur in the CC. Fast inference is possible by linear programming relaxation, and effective parameter learning by structured support vector machine is also possible by incorporating a decomposable structured loss function. Experimental results on various data sets show that the proposed HO-CC outperforms other state-of-the-art image segmentation algorithms. The HO-CC framework is therefore an efficient and flexible image segmentation framework.

This paper makes two contributions: the first is the proposal of a new model-The associative hierarchical random field (AHRF), and a novel algorithm for its optimization; the second is the application of this model to the problem of semantic segmentation. Most methods for semantic segmentation are formulated as a labeling problem for variables that might correspond to either pixels or segments such as super-pixels. It is well known that the generation of super pixel segmentations is not unique. This has motivated many researchers to use multiple super pixel segmentations for problems such as semantic segmentation or single view reconstruction. These super-pixels have not yet been combined in a principled manner, this is a difficult problem, as they may overlap, or be nested in such a way that the segmentations form a segmentation tree. Our new hierarchical random field model allows information from all of the multiple segmentations to contribute to a global energy. MAP inference in this model can be performed efficiently using powerful graph cut based move making algorithms. Our framework generalizes much of the previous work based on pixels or segments, and the resulting labelings can be viewed both as a detailed segmentation at the pixel level, or at the other extreme, as a segment selector that pieces together a solution like a jigsaw, selecting the best segments from different segmentations as pieces. We evaluate its performance on some of the most challenging data sets for object class segmentation, and show that this ability to perform inference using multiple overlapping segmentations leads to state-of-the-art results.

Energy minimization algorithms, such as graph cuts, enable the computation of the MAP solution under certain probabilistic models such as Markov random fields. However, for many computer vision problems, the MAP solution under the model is not the ground truth solution. In many problem scenarios, the system has access to certain statistics of the ground truth. For instance, in image segmentation, the area and boundary length of the object may be known. In these cases, we want to estimate the most probable solution that is consistent with such statistics, i.e., satisfies certain equality or inequality constraints. The above constrained energy minimization problem is NP-hard in general, and is usually solved using Linear Programming formulations, which relax the integrality constraints. This paper proposes a novel method that directly finds the discrete approximate solution of such problems by maximizing the corresponding Lagrangian dual. This method can be applied to any constrained energy minimization problem whose unconstrained version is polynomial time solvable, and can handle multiple, equality or inequality, and linear or non-linear constraints. One important advantage of our method is the ability to handle second order constraints with both-side inequalities with a weak restriction, not trivial in the relaxation based methods, and show that the restriction does not affect the accuracy in our cases.We demonstrate the efficacy of our method on the foreground/background image segmentation problem, and show that it produces impressive segmentation results with less error, and runs more than 20 times faster than the state-of-the-art LP relaxation based approaches.

In the last few years, substantially different approaches have been adopted for segmenting and detecting "things" (object categories that have a well defined shape such as people and cars) and "stuff" (object categories which have an amorphous spatial extent such as grass and sky). While things have been typically detected by sliding window or Hough transform based methods, detection of stuff is generally formulated as a pixel or segment-wise classification problem. This paper proposes a framework for scene understanding that models both things and stuff using a common representation while preserving their distinct nature by using a property list. This representation allows us to enforce sophisticated geometric and semantic relationships between thing and stuff categories via property interactions in a single graphical model. We use the latest advances made in the field of discrete optimization to efficiently perform maximum a posteriori (MAP) inference in this model. We evaluate our method on the Stanford dataset by comparing it against state-of-the-art methods for object segmentation and detection. We also show that our method achieves competitive performances on the challenging PASCAL '09 segmentation dataset.

We present a machine learning technique for estimating absolute, per-pixel depth using any conventional monocular 2D camera, with minor hardware modifications. Our approach targets close-range human capture and interaction where dense 3D estimation of hands and faces is desired. We use hybrid classification-regression forests to learn how to map from near infrared intensity images to absolute , metric depth in real-time. We demonstrate a variety of human-computer interaction and capture scenarios. Experiments show an accuracy that outperforms a conventional light fall-off baseline, and is comparable to high-quality consumer depth cameras, but with a dramatically reduced cost, power consumption, and form-factor.

Representation languages for coalitional games are a key research area in algorithmic game theory. There is an inherent tradeoff between how general a language is, allowing it to capture more elaborate games, and how hard it is computationally to optimize and solve such games. One prominent such language is the simple yet expressive Weighted Graph Games (WGGs) representation (Deng and Papadimitriou, 1994), which maintains knowledge about synergies between agents in the form of an edge weighted graph. We consider the problem of finding the optimal coalition structure in WGGs. The agents in such games are vertices in a graph, and the value of a coalition is the sum of the weights of the edges present between coalition members. The optimal coalition structure is a partition of the agents to coalitions, that maximizes the sum of utilities obtained by the coalitions. We show that finding the optimal coalition structure is not only hard for general graphs, but is also intractable for restricted families such as planar graphs which are amenable for many other combinatorial problems. We then provide algorithms with constant factor approximations for planar, minor-free and bounded degree graphs.

We study recommendation in scenarios where there's no prior information about the quality of content in the system. We present an online algorithm that continually optimizes recommendation relevance based on behavior of past users. Our method trades weaker theoretical guarantees in asymptotic performance than the state-of-the-art for stronger theoretical guarantees in the online setting. We test our algorithm on real-world data collected from previous recommender systems and show that our algorithm learns faster than existing methods and performs equally well in the long-run.

Training of Structural SVMs involves solving a large Quadratic Program (QP). One popular method for solving this QP is a cutting-plane approach, where the most violated constraint is iteratively added to a working-set of constraints. Unfortunately, training models with a large number of parameters remains a time consuming process. This paper shows that significant computational savings can be achieved by adding multiple diverse and highly violated constraints at every iteration of the cutting-plane algorithm. We show that generation of such diverse cutting-planes involves extracting diverse M-Best solutions from the loss-augmented score of the training instances. To find these diverse M-Best solutions, we employ a recently proposed algorithm [4]. Our experiments on image segmentation and protein side-chain prediction show that the proposed approach can lead to significant computational savings, e.g., ∼28% reduction in training time.

We discuss a model for image segmentation that is able to overcome the short-boundary bias observed in standard pairwise random field based approaches. To wit, we show that a random field with multi-layered hidden units can encode boundary preserving higher order potentials such as the ones used in the cooperative cuts model of [11] while still allowing for fast and exact MAP inference. Exact inference allows our model to outperform previous image segmentation methods, and to see the true effect of coupling graph edges. Finally, our model can be easily extended to handle segmentation instances with multiple labels, for which it yields promising results.

Conventional decision forest based methods for image labelling tasks like object segmentation make predictions for each variable (pixel) independently [3, 5, 8]. This prevents them from enforcing dependencies between variables and translates into locally inconsistent pixel labellings. Random field models, instead, encourage spatial consistency of labels at increased computational expense. This paper presents a new and efficient forest based model that achieves spatially consistent semantic image segmentation by encoding variable dependencies directly in the feature space the forests operate on. Such correlations are captured via new long-range, soft connectivity features, computed via generalized geodesic distance transforms. Our model can be thought of as a generalization of the successful Semantic Texton Forest, Auto-Context, and Entangled Forest models. A second contribution is to show the connection between the typical Conditional Random Field (CRF) energy and the forest training objective. This analysis yields a new objective for training decision forests that encourages more accurate structured prediction. Our GeoF model is validated quantitatively on the task of semantic image segmentation, on four challenging and very diverse image datasets. GeoF outperforms both state of-the-art forest models and the conventional pair wise CRF.

Traditional video compression methods obtain a compact representation for image frames by computing coarse motion fields defined on patches of pixels called blocks, in order to compensate for the motion in the scene across frames. This piecewise constant approximation makes the motion field efficiently encodable, but it introduces block artifacts in the warped image frame. In this paper, we address the problem of estimating dense motion fields that, while accurately predicting one frame from a given reference frame by warping it with the field, are also compressible. We introduce a representation for motion fields based on wavelet bases, and approximate the compressibility of their coefficients with a piecewise smooth surrogate function that yields an objective function similar to classical optical flow formulations. We then show how to quantize and encode such coefficients with adaptive precision. We demonstrate the effectiveness of our approach by comparing its performance with a state-of-the-art wavelet video encoder. Experimental results on a number of standard flow and video datasets reveal that our method significantly outperforms both block-based and optical-flow-based motion compensation algorithms.

This paper addresses the problem of semantic segmentation of 3D point clouds. We extend the inference machines framework of Ross et al. by adding spatial factors that model mid-range and long-range dependencies inherent in the data. The new model is able to account for semantic spatial context. During training, our method automatically isolates and retains factors modelling spatial dependencies between variables that are relevant for achieving higher prediction accuracy. We evaluate the proposed method by using it to predict 17-category semantic segmentations on sets of stitched Kinect scans. Experimental results show that the spatial dependencies learned by our method significantly improve the accuracy of segmentation. They also show that our method outperforms the existing segmentation technique of Koppula et al.

Scene understanding is an important yet very challenging problem in computer vision. In the past few years, researchers have taken advantage of the recent diffusion of depth-RGB (RGB-D) cameras to help simplify the problem of inferring scene semantics. However, while the added 3D geometry is certainly useful to segment out objects with different depth values, it also adds complications in that the 3D geometry is often incorrect because of noisy depth measurements and the actual 3D extent of the objects is usually unknown because of occlusions. In this paper we propose a new method that allows us to jointly refine the 3D reconstruction of the scene (raw depth values) while accurately segmenting out the objects or scene elements from the 3D reconstruction. This is achieved by introducing a new model which we called Voxel-CRF. The Voxel-CRF model is based on the idea of constructing a conditional random field over a 3D volume of interest which captures the semantic and 3D geometric relationships among different elements (voxels) of the scene. Such model allows to jointly estimate (1) a dense voxel-based 3D reconstruction and (2) the semantic labels associated with each voxel even in presence of partial occlusions using an approximate yet efficient inference strategy. We evaluated our method on the challenging NYU Depth dataset (Version 1 and 2). Experimental results show that our method achieves competitive accuracy in inferring scene semantics and visually appealing results in improving the quality of the 3D reconstruction. We also demonstrate an interesting application of object removal and scene completion from RGB-D images.

While modern off-the-shelf OCR engines show particularly high accuracy on scanned text, text detection and recognition in natural images still remains a challenging problem. Here, we demonstrate that OCR engines can still perform well on this harder task as long as appropriate image binarization is applied to input photographs. For such binarization, we systematically evaluate the performance of 12 binarization methods as well as of a new binarization algorithm that we propose here. Our evaluation includes different metrics and uses established natural image text recognition benchmarks (ICDAR 2003 and ICDAR 2011). Our main finding is thus the fact that image binarization methods combined with additional filtering of generated connected components and off-the-shelf OCR engines can achieve state-of-the-art performance for end-to-end text understanding in natural images.

We introduce new approaches for augmenting annotated training datasets used for object detection tasks that serve achieving two goals: reduce the effort needed for collecting and manually annotating huge datasets and introduce novel variations to the initial dataset that help the learning algorithms. The methods presented in this work aim at relocating objects using their segmentation masks to new backgrounds. These variations comprise changes in properties of objects such as spatial location in the image, surrounding context and scale. We propose a model selection approach to arbitrate between the constructed model on a per class basis. Experimental results show gains that can be harvested using the proposed approach.

Randomized decision trees and forests have a rich history in machine learning and have seen considerable success in application, perhaps particularly so for computer vision. However, they face a fundamental limitation: given enough data, the number of nodes in decision trees will grow exponentially with depth. For certain applications, for example on mobile or embedded processors, memory is a limited resource, and so the exponential growth of trees limits their depth, and thus their potential accuracy. This paper proposes decision jungles, revisiting the idea of ensembles of rooted decision directed acyclic graphs (DAGs), and shows these to be compact and powerful discriminative models for classification. Unlike conventional decision trees that only allow one path to every node, a DAG in a decision jungle allows multiple paths from the root to each leaf. We present and compare two new node merging algorithms that jointly optimize both the features and the structure of the DAGs efficiently. During training, node splitting and node merging are driven by the minimization of exactly the same objective function, here the weighted sum of entropies at the leaves. Results on varied datasets show that, compared to decision forests and several other baselines, decision jungles require dramatically less memory while considerably improving generalization.

We describe two new approaches to human pose estimation. Both can quickly and accurately predict the 3D positions of body joints from a single depth image without using any temporal information. The key to both approaches is the use of a large, realistic, and highly varied synthetic set of training images. This allows us to learn models that are largely invariant to factors such as pose, body shape, field-of-view cropping, and clothing. Our first approach employs an intermediate body parts representation, designed so that an accurate per-pixel classification of the parts will localize the joints of the body. The second approach instead directly regresses the positions of body joints. By using simple depth pixel comparison features and parallelizable decision forests, both approaches can run super-real time on consumer hardware. Our evaluation investigates many aspects of our methods, and compares the approaches to each other and to the state of the art. Results on silhouettes suggest broader applicability to other imaging modalities.

Image partitioning is an important preprocessing step for many of the state-of-the-art algorithms used for performing high-level computer vision tasks. Typically, partitioning is conducted without regard to the task in hand. We propose a task-specific image partitioning framework to produce a region-based image representation that will lead to a higher task performance than that reached using any task-oblivious partitioning framework and existing supervised partitioning framework, albeit few in number. The proposed method partitions the image by means of correlation clustering, maximizing a linear discriminant function defined over a superpixel graph. The parameters of the discriminant function that define task-specific similarity/dissimilarity among superpixels are estimated based on structured support vector machine (S-SVM) using task-specific training data. The S-SVM learning leads to a better generalization ability while the construction of the superpixel graph used to define the discriminant function allows a rich set of features to be incorporated to improve discriminability and robustness. We evaluate the learned task-aware partitioning algorithms on three benchmark datasets. Results show that task-aware partitioning leads to better labeling performance than the partitioning computed by the state-of-the-art general-purpose and supervised partitioning algorithms. We believe that the task-specific image partitioning paradigm is widely applicable to improving performance in high-level image understanding tasks.

Entertainment and gaming systems such as the Wii and XBox Kinect have brought touchless, body-movement based interfaces to the masses. Systems like these enable the estimation of movements of various body parts from raw inertial motion or depth sensor data. However, the interface developer is still left with the challenging task of creating a system that recognizes these movements as embodying meaning. The machine learning approach for tackling this problem requires the collection of data sets that contain the relevant body movements and their associated semantic labels. These data sets directly impact the accuracy and performance of the gesture recognition system and should ideally contain all natural variations of the movements associated with a gesture. This paper addresses the problem of collecting such gesture datasets. In particular, we investigate the question of what is the most appropriate semiotic modality of instructions for conveying to human subjects the movements the system developer needs them to perform. The results of our qualitative and quantitative analysis indicate that the choice of modality has a significant impact on the performance of the learnt gesture recognition system; particularly in terms of correctness and coverage.

Random forests have been successfully applied to various high level computer vision tasks such as human pose estimation and object segmentation. These models are extremely efficient but work under the assumption that the output variables (such as body part locations or pixel labels) are independent. In this paper, we present a conditional regression forest model for human pose estimation that incorporates dependency relationships between output variables through a global latent variable while still maintaining a low computational cost. We show that the incorporation of a global latent variable encoding torso orientation, or human height, etc., can dramatically increase the accuracy of body joint location prediction. Our model also allows efficient and seamless incorporation of prior knowledge about the problem instance such as the height or orientation of the human subject which can be available from the problem context or via a temporal model. We show that our method significantly outperforms state-of-the-art methods for pose estimation from depth images. The conditional regression model proposed in the paper is general and can be applied to other problems where random forests are used.

In this paper we propose a new graph formulation for solving Discrete Tomography problems in the case where only a very few number of projections are available. Graph formulations are efficient to solve many different pixel labeling problems in Image Processing. However, applying graph models for Discrete Tomography problems is a very challenging task due to the high dimensionality of the data and the complexity of the constraints formulations. Due to the NP-hardness of the problem, even the computation of a local minima is still a challenge. In this paper, we propose a graph model with a polynomial number of additional edges formulating the projection consistency and an iterative algorithm to efficiently minimize the energy function. Our formulation aims to provide 3D results based on a restricted number of projections.

The paper addresses the problem of generating multiple hypotheses for prediction tasks that involve interaction with users or successive components in a cascade. Given a set of multiple hypotheses, such components/users have the ability to automatically rank the results and thus retrieve the best one. The standard approach for handling this scenario is to learn a single model and then produce M-best Maximum a Posteriori (MAP) hypotheses from this model. In contrast, we formulate this multiple {\em choice} learning task as a multiple-output structured-output prediction problem with a loss function that captures the natural setup of the problem. We present a max-margin formulation that minimizes an upper-bound on this loss-function. Experimental results on the problems of image co-segmentation and protein side-chain prediction show that our method outperforms conventional approaches used for this scenario and leads to substantial improvements in prediction accuracy.

In this paper we introduce Context-Sensitive Decision Forests - A new perspective to exploit contextual information in the popular decision forest framework for the object detection problem. They are tree-structured classifiers with the ability to access intermediate prediction (here: classification and regression) information during training and inference time. This intermediate prediction is available to each sample, which allows us to develop context-based decision criteria, used for refining the prediction process. In addition, we introduce a novel split criterion which in combination with a priority based way of constructing the trees, allows more accurate regression mode selection and hence improves the current context information. In our experiments, we demonstrate improved results for the task of pedestrian detection on the challenging TUD data set when compared to state-of-the-art methods.

Abstract Optical flow is a critical component of video editing applications, e.g. for tasks such as object tracking, segmentation, and selection. In this paper, we propose an optical flow algorithm calledSimpleFlowwhose running times increase sublinearly in the number of pixels. Central to our approach is a probabilistic representation of the motion flow that is computed using only local evidence and without resorting to global optimization. To estimate the flow in image regions where the motion is smooth, we use a sparse set of samples only, thereby avoiding the expensive computation inherent in traditional dense algorithms. We show that our results can be used as is for a variety of video editing tasks. For applications where accuracy is paramount, we use our result to bootstrap a global optimization. This significantly reduces the running times of such methods without sacrificing accuracy. We also demonstrate that the SimpleFlow algorithm can process HD and 4K footage in reasonable times.

Hough transform-based methods for detecting multiple objects use nonmaxima suppression or mode seeking to locate and distinguish peaks in Hough images. Such postprocessing requires the tuning of many parameters and is often fragile, especially when objects are located spatially close to each other. In this paper, we develop a new probabilistic framework for object detection which is related to the Hough transform. It shares the simplicity and wide applicability of the Hough transform but, at the same time, bypasses the problem of multiple peak identification in Hough images and permits detection of multiple objects without invoking nonmaximum suppression heuristics. Our experiments demonstrate that this method results in a significant improvement in detection accuracy both for the classical task of straight line detection and for a more modern category-level (pedestrian) detection problem.

Image completion is an important photo-editing task which involves synthetically filling a hole in the image such that the image still appears natural. State-of-the-art image completion methods work by searching for patches in the image that fit well in the hole region. Our key insight is that image patches remain natural under a variety of transformations (such as scale, rotation and brightness change), and it is important to exploit this. We propose and investigate the use of different optimisation methods to search for the best patches and their respective transformations for producing consistent, improved completions. Experiments on a number of challenging problem instances demonstrate that our methods outperform state-of-the-art techniques. © 2011. The copyright of this document resides with its authors.

This paper presents a new adaptive graph-cut based move-making algorithm for energy minimization. Traditional move-making algorithms such as Expansion and Swap operate by searching for better solutions in some predefined moves spaces around the current solution. In contrast, our algorithm uses the primal-dual interpretation of the Expansion-move algorithm to adaptively compute the best move-space to search over. At each step, it tries to greedily find the move-space that will lead to biggest decrease in the primal-dual gap. We test different variants of our algorithm on a variety of image labelling problems such as object segmentation and stereo. Experimental results show that our adaptive strategy significantly outperforms the conventional Expansion-move algorithm, in some cases cutting the runtime by 50%.

This paper presents a method for joint stereo matching and object segmentation. In our approach a 3D scene is represented as a collection of visually distinct and spatially coherent objects. Each object is characterized by three different aspects: a color model, a 3D plane that approximates the object's disparity distribution, and a novel 3D connectivity property. Inspired by Markov Random Field models of image segmentation, we employ object-level color models as a soft constraint, which can aid depth estimation in powerful ways. In particular, our method is able to recover the depth of regions that are fully occluded in one input view, which to our knowledge is new for stereo matching. Our model is formulated as an energy function that is optimized via fusion moves. We show high-quality disparity and object segmentation results on challenging image pairs as well as standard benchmarks. We believe our work not only demonstrates a novel synergy between the areas of image segmentation and stereo matching, but may also inspire new work in the domain of automatic and interactive object-level scene manipulation.

This paper addresses the problem of efficiently solving large-scale energy minimization problems encountered in computer vision. We propose an energy-aware method for merging random variables to reduce the size of the energy to be minimized. The method examines the energy function to find groups of variables which are likely to take the same label in the minimum energy state and thus can be represented by a single random variable. We propose and evaluate a number of extremely efficient variable grouping strategies. Experimental results show that our methods result in a dramatic reduction in the computational cost and memory requirements (in some cases by a factor of one hundred) with almost no drop in the accuracy of the final result. Comparative evaluation with efficient super-pixel generation methods, which are commonly used in variable grouping, reveals that our methods are far superior both in terms of accuracy and running time.

We present a new approach to general-activity human pose estimation from depth images, building on Hough forests. We extend existing techniques in several ways: real time prediction of multiple 3D joints, explicit learning of voting weights, vote compression to allow larger training sets, and a comparison of several decision-tree training objectives. Key aspects of our work include: regression directly from the raw depth image, without the use of an arbitrary intermediate representation; applicability to general motions (not constrained to particular activities) and the ability to localize occluded as well as visible body joints. Experimental results demonstrate that our method produces state of the art results on several data sets including the challenging MSRC-5000 pose estimation test set, at a speed of about 200 frames per second. Results on silhouettes suggest broader applicability to other imaging modalities.

This paper introduces a new formulation for discrete image labeling tasks, the Decision Tree Field (DTF), that combines and generalizes random forests and conditional random fields (CRF) which have been widely used in computer vision. In a typical CRF model the unary potentials are derived from sophisticated random forest or boosting based classifiers, however, the pairwise potentials are assumed to (1) have a simple parametric form with a pre-specified and fixed dependence on the image data, and (2) to be defined on the basis of a small and fixed neighborhood. In contrast, in DTF, local interactions between multiple variables are determined by means of decision trees evaluated on the image data, allowing the interactions to be adapted to the image content. This results in powerful graphical models which are able to represent complex label structure. Our key technical contribution is to show that the DTF model can be trained efficiently and jointly using a convex approximate likelihood function, enabling us to learn over a million free model parameters. We show experimentally that for applications which have a rich and complex label structure, our model achieves excellent results.

We present a system for accurate real-time mapping of complex and arbitrary indoor scenes in variable lighting conditions, using only a moving low-cost depth camera and commodity graphics hardware. We fuse all of the depth data streamed from a Kinect sensor into a single global implicit surface model of the observed scene in real-time. The current sensor pose is simultaneously obtained by tracking the live depth frame relative to the global model using a coarse-to-fine iterative closest point (ICP) algorithm, which uses all of the observed depth data available. We demonstrate the advantages of tracking against the growing full surface model compared with frame-to-frame tracking, obtaining tracking and mapping results in constant time within room sized scenes with limited drift and high accuracy. We also show both qualitative and quantitative results relating to various aspects of our tracking and mapping system. Modelling of natural scenes, in real-time with only commodity sensor and GPU hardware, promises an exciting step forward in augmented reality (AR), in particular, it allows dense surfaces to be reconstructed in real-time, with a level of detail and robustness beyond any solution yet presented using passive computer vision.

For many of the state-of-the-art computer vision algorithms, image segmentation is an important preprocessing step. As such, several image segmentation algorithms have been proposed, however, with certain reservation due to high computational load and many hand-tuning parameters. Correlation clustering, a graph-partitioning algorithm often used in natural language processing and document clustering, has the potential to perform better than previously proposed image segmentation algorithms. We improve the basic correlation clustering formulation by taking into account higher-order cluster relationships. This improves clustering in the presence of local boundary ambiguities. We first apply the pairwise correlation clustering to image segmentation over a pairwise superpixel graph and then develop higher-order correlation clustering over a hypergraph that considers higher-order relations among superpixels. Fast inference is possible by linear programming relaxation, and also effective parameter learning framework by structured support vector machine is possible. Experimental results on various datasets show that the proposed higher-order correlation clustering outperforms other state-of-the-art image segmentation algorithms.

KinectFusion enables a user holding and moving a standard Kinect camera to rapidly create detailed 3D reconstructions of an indoor scene. Only the depth data from Kinect is used to track the 3D pose of the sensor and reconstruct, geometrically precise, 3D models of the physical scene in real-time. The capabilities of KinectFusion, as well as the novel GPU-based pipeline are described in full. Uses of the core system for low-cost handheld scanning, and geometry-aware augmented reality and physics-based interactions are shown. Novel extensions to the core GPU pipeline demonstrate object segmentation and user interaction directly in front of the sensor, without degrading camera tracking or reconstruction. These extensions are used to enable real-time multi-touch interactions anywhere, allowing any planar or non-planar reconstructed physical surface to be appropriated for touch.

We consider optimizing the coalition structure in Coalitional Skill Games (CSGs), a succinct representation of coalitional games. In CSGs, the value of a coalition depends on the tasks its members can achieve. The tasks require various skills to complete them, and agents may have different skill sets. The optimal coalition structure is a partition of the agents to coalitions, that maximizes the sum of utilities obtained by the coalitions. We show that CSGs can represent any characteristic function, and consider optimal coalition structure generation in this representation. We provide hardness results, showing that in general CSGs, as well as in very restricted versions of them, computing the optimal coalition structure is hard. On the positive side, we show that the problem can be reformulated as constraint satisfaction on a hyper graph, and present an algorithm that finds the optimal coalition structure in polynomial time for instances with bounded tree-width and number of tasks.

To detect multiple objects of interest, the methods based on Hough transform use non-maxima supression or mode seeking in order to locate and to distinguish peaks in Hough images. Such postprocessing requires tuning of extra parameters and is often fragile, especially when objects of interest tend to be closely located. In the paper, we develop a new probabilistic framework that is in many ways related to Hough transform, sharing its simplicity and wide applicability. At the same time, the framework bypasses the problem of multiple peaks identification in Hough images, and permits detection of multiple objects without invoking nonmaximum suppression heuristics. As a result, the experiments demonstrate a significant improvement in detection accuracy both for the classical task of straight line detection and for a more modern category-level (pedestrian) detection problem.

This paper proposes a new stereo model which encodes the simple assumption that the scene is composed of a few, smooth surfaces. A key feature of our model is the surface-based representation, where each pixel is assigned to a 3D surface (planes or B-splines). This representation enables several important contributions: Firstly, we formulate a higher-order prior which states that pixels of similar appearance are likely to belong to the same 3D surface. This enables to incorporate the very popular color segmentation constraint in a soft and principled way. Secondly, we use a global MDL prior to penalize the number of surfaces. Thirdly, we are able to incorporate, in a simple way, a prior which favors low curvature surfaces. Fourthly, we improve the asymmetric occlusion model by disallowing pixels of the same surface to occlude each other. Finally, we use the known fusion move approach which enables a powerful optimization of our model, despite the infinite number of possible labelings (surfaces).

Markov random fields with higher order potentials have emerged as a powerful model for several problems in computer vision. In order to facilitate their use, we propose a new representation for higher order potentials as upper and lower envelopes of linear functions. Our representation concisely models several commonly used higher order potentials, thereby providing a unified framework for minimizing the corresponding Gibbs energy functions. We exploit this framework by converting lower envelope potentials to standard pairwise functions with the addition of a small number of auxiliary variables. This allows us to minimize energy functions with lower envelope potentials using conventional algorithms such as BP, TRW and α-expansion. Furthermore, we show how the minimization of energy functions with upper envelope potentials leads to a difficult minmax problem. We address this difficulty by proposing a new message passing algorithm that solves a linear programming relaxation of the problem. Although this is primarily a theoretical paper, we demonstrate the efficacy of our approach on the binary (fg/bg) segmentation problem.

In this paper we considerably improve on a state-of-the-art alpha matting approach by incorporating a new prior which is based on the image formation process. In particular, we model the prior probability of an alpha matte as the convolution of a high-resolution binary segmentation with the spatially varying point spread function (PSF) of the camera. Our main contribution is a new and efficient de-convolution approach that recovers the prior model, given an approximate alpha matte. By assuming that the PSF is a kernel with a single peak, we are able to recover the binary segmentation with an MRF-based approach, which exploits flux and a new way of enforcing connectivity. The spatially varying PSF is obtained via a partitioning of the image into regions of similar defocus. Incorporating our new prior model into a state-of-the-art matting technique produces results that outperform all competitors, which we confirm using a publicly available benchmark.

In this paper, we present novel techniques that improve the computational and memory efficiency of algorithms for solving multilabel energy functions arising from discrete mrfs or crfs. These methods are motivated by the observations that the performance of minimization algorithms depends on: 1) the initialization used for the primal and dual variables and 2) the number of primal variables involved in the energy function. Our first method (dynamic alpha-expansion) works by "recycling" results from previous problem instances. The second method simplifies the energy function by "reducing" the number of unknown variables present in the problem. Further, we show that it can also be used to generate a good initialization for the dynamic alpha-expansion algorithm by "reusing" dual variables. We test the performance of our methods on energy functions encountered in the problems of stereo matching and color and object-based segmentation. Experimental results show that our methods achieve a substantial improvement in the performance of alpha-expansion, as well as other popular algorithms such as sequential tree-reweighted message passing and max-product belief propagation. We also demonstrate the applicability of our schemes for certain higher order energy functions, such as the one described in [1], for interactive texture-based image and video segmentation. In most cases, we achieve a 10-15 times speed-up in the computation time. Our modified alpha-expansion algorithm provides similar performance to Fast-PD, but is conceptually much simpler. Both alpha-expansion and Fast-PD can be made orders of magnitude faster when used in conjunction with the "reduce" scheme proposed in this paper.

The availability of quantitative online benchmarks for low-level vision tasks such as stereo and optical flow has led to significant progress in the respective fields. This paper introduces such a benchmark for image matting. There are three key factors for a successful benchmarking system: (a) a challenging, high-quality ground truth test set; (b) an online evaluation repository that is dynamically updated with new results; (c) perceptually motivated error functions. Our new benchmark strives to meet all three criteria. We evaluated several matting methods with our benchmark and show that their performance varies depending on the error function. Also, our challenging test set reveals problems of existing algorithms, not reflected in previously reported results. We hope that our effort will lead to considerable progress in the field of image matting, and welcome the reader to visit our benchmark at www.aIphamatting.com.

Higher order energy functions have the ability to encode high level structural dependencies between pixels, which have been shown to be extremely powerful for image labeling problems. Their use, however, is severely hampered in practice by the intractable complexity of representing and minimizing such functions. We observed that higher order functions encountered in computer vision are very often “sparse”, i.e. many labelings of a higher order clique are equally unlikely and hence have the same high cost. In this paper, we address the problem of minimizing such sparse higher order energy functions. Our method works by transforming the problem into an equivalent quadratic function minimization problem. The resulting quadratic function can be minimized using popular message passing or graph cut based algorithms for MAP inference. Although this is primarily a theoretical paper, it also shows how higher order functions can be used to obtain impressive results for the binary texture restoration problem.

Most methods for object class segmentation are formulated as a labelling problem over a single choice of quantisation of an image space - pixels, segments or group of segments. It is well known that each quantisation has its fair share of pros and cons; and the existence of a common optimal quantisation level suitable for all object categories is highly unlikely. Motivated by this observation, we propose a hierarchical random field model, that allows integration of features computed at different levels of the quantisation hierarchy. MAP inference in this model can be performed efficiently using powerful graph cut based move making algorithms. Our framework generalises much of the previous work based on pixels or segments. We evaluate its efficiency on some of the most challenging data-sets for object class segmentation, and show it obtains state-of-the-art results.

User-provided object bounding box is a simple and popular interaction paradigm considered by many existing interactive image segmentation frameworks. However, these frameworks tend to exploit the provided bounding box merely to exclude its exterior from consideration and sometimes to initialize the energy minimization. In this paper, we discuss how the bounding box can be further used to impose a powerful topological prior, which prevents the solution from excessive shrinking and ensures that the user-provided box bounds the segmentation in a sufficiently tight way. The prior is expressed using hard constraints incorporated into the global energy minimization framework leading to an NP-hard integer program. We then investigate the possible optimization strategies including linear relaxation as well as a new graph cut algorithm called pinpointing. The latter can be used either as a rounding method for the fractional LP solution, which is provably better than thresholding-based rounding, or as a fast standalone heuristic. We evaluate the proposed algorithms on a publicly available dataset, and demonstrate the practical benefits of the new prior both qualitatively and quantitatively.

We consider the question of computing Maximum A Posteriori (MAP) assignment in an arbitrary pair-wise Markov Random Field (MRF). We present a randomized iterative algorithm based on simple local updates. The algorithm, starting with an arbitrary initial assignment, updates it in each iteration by first, picking a random node, then selecting an (appropriately chosen) random local neighborhood and optimizing over this local neighborhood. Somewhat surprisingly, we show that this algorithm finds a near optimal assignment within $2n\ln n$ iterations on average and with high probability for {\em any} $n$ node pair-wise MRF with {\em geometry} (i.e. MRF graph with polynomial growth) with the approximation error depending on (in a reasonable manner) the geometric growth rate of the graph and the average radius of the local neighborhood -- this allows for a graceful tradeoff between the complexity of the algorithm and the approximation error. Through extensive simulations, we show that our algorithm finds extremely good approximate solutions for various kinds of MRFs with geometry.

In this paper, we extend the class of energy functions for which the optimal \alpha-expansion and \alpha \beta-swap moves can be computed in polynomial time. Specifically, we introduce a novel family of higher order clique potentials, and show that the expansion and swap moves for any energy function composed of these potentials can be found by minimizing a submodular function. We also show that for a subset of these potentials, the optimal move can be found by solving an st-mincut problem. We refer to this subset as the {\cal P};n Potts model. Our results enable the use of powerful \alpha-expansion and \alpha \beta-swap move making algorithms for minimization of energy functions involving higher order cliques. Such functions have the capability of modeling the rich statistics of natural scenes and can be used for many applications in Computer Vision. We demonstrate their use in one such application, i.e., the texture-based image or video-segmentation problem.

In this paper, we present novel techniques that improve the computational and memory efficiency of algorithms for solving multi-label energy functions arising from discrete MRFs or CRFs. These methods are motivated by the observations that the performance of minimization algorithms depends on: (a) the initialization used for the primal and dual variables; and (b) the number of primal variables involved in the energy function. Our first method (dynamic alpha-expansion) works by dasiarecyclingpsila results from previous problem instances. The second method simplifies the energy function by dasiareducingpsila the number of unknown variables, and can also be used to generate a good initialization for the dynamic alpha-expansion algorithm by dasiareusingpsila dual variables. We test the performance of our methods on energy functions encountered in the problems of stereo matching, and colour and object based segmentation. Experimental results show that our methods achieve a substantial improvement in the performance of alpha-expansion, as well as other popular algorithms such as sequential tree-reweighted message passing, and max-product belief propagation. In most cases we achieve a 10-15 times speed-up in the computation time. Our modified alpha-expansion algorithm provides similar performance to Fast-PD. However, it is much simpler and can be made orders of magnitude faster by using the initialization schemes proposed in the paper.

This paper proposes a novel framework for labelling problems which is able to combine multiple segmentations in a principled manner. Our method is based on higher order conditional random fields and uses potentials defined on sets of pixels (image segments) generated using unsupervised segmentation algorithms. These potentials enforce label consistency in image regions and can be seen as a strict generalization of the commonly used pairwise contrast sensitive smoothness potentials. The higher order potential functions used in our framework take the form of the Robust Pnmodel. This enables the use of powerful graph cut based move making algorithms for performing inference in the framework [14]. We test our method on the problem of multi-class object segmentation by augmenting the conventional CRF used for object segmentation with higher order potentials defined on image regions. Experiments on challenging data sets show that integration of higher order potentials quantitatively and qualitatively improves results leading to much better definition of object boundaries. We believe that this method can be used to yield similar improvements for many other labelling problems.

This paper addresses the problem of exactly inferring the maximum a posteriori solutions of discrete multi-label MRFs or CRFs with higher order cliques. We present a framework to transform special classes of multi-label higher order functions to submodular second order Boolean functions (referred to as Fs2), which can be minimized exactly using graph cuts and we characterize those classes. The basic idea is to use two or more Boolean variables to encode the states of a single multi-label variable. There are many ways in which this can be done and much interesting research lies in finding ways which are optimal or minimal in some sense. We study the space of possible encodings and find the ones that can transform the most general class of functions to Fs2. Our main contributions are two-fold. First, we extend the subclass of submodular energy functions that can be minimized exactly using graph cuts. Second, we show how higher order potentials can be used to improve single view 3D reconstruction results. We believe that our work on exact minimization of higher order energy functions will lead to similar improvements in solutions of other labelling problems.

We consider the problem of optimizing multilabel MRFs, which is in general NP-hard and ubiquitous in low-level computer vision. One approach for its solution is to formulate it as an integer linear programming and relax the integrality constraints. The approach we consider in this paper is to first convert the multi-label MRF into an equivalent binary-label MRF and then to relax it. The resulting relaxation can be efficiently solved using a maximum flow algorithm. Its solution provides us with a partially optimal labelling of the binary variables. This partial labelling is then easily transferred to the multi-label problem. We study the theoretical properties of the new relaxation and compare it with the standard one. Specifically, we compare tightness, and characterize a subclass of problems where the two relaxations coincide. We propose several combined algorithms based on the technique and demonstrate their performance on challenging computer vision problems.

We introduce a new representation for video which facilitates a number of common editing tasks. The representation has some of the power of a full reconstruction of 3D surface models from video, but is designed to be easy to recover from a priori unseen and uncalibrated footage. By modelling the image-formation process as a 2D-to-2D transformation from an object's texture map to the image, modulated by an object-space occlusion mask, we can recover a representation which we term the "unwrap mosaic". Many editing operations can be performed on the unwrap mosaic, and then re-composited into the original sequence, for example resizing objects, repainting textures, copying/cutting/pasting objects, and attaching effects layers to deforming objects.

In this paper we extend the class of energy functions for which the optimal alpha-expansion and alphabeta-swap moves can be computed in polynomial time. Specifically, we introduce a class of higher order clique potentials and show that the expansion and swap moves for any energy function composed of these potentials can be found by minimizing a submodular function. We also show that for a subset of these potentials, the optimal move can be found by solving an st-mincut problem. We refer to this subset as the P3Potts model. Our results enable the use of powerful move making algorithms i.e. alpha-expansion and alphabeta-swap for minimization of energy functions involving higher order cliques. Such functions have the capability of modelling the rich statistics of natural scenes and can be used for many applications in computer vision. We demonstrate their use on one such application i.e. the texture based video segmentation problem.

Abstract-In this paper we present a fast new fully dynamic algorithm for the st-mincut/max-flow problem. We show how this algorithm can be used to efficiently compute MAP solutions for certain dynamically changing MRF models in computer vision such as image segmentation. Specifically, given the solution of the max-flow problem on a graph, the dynamic algorithm efficiently computes the maximum flow in a modified version of the graph. The time taken by it is roughly proportional to the total amount of change in the edge weights of the graph. Our experiments show that, when the number of changes in the graph is small, the dynamic algorithm is significantly faster than the best known static graph cut algorithm. We test the performance of our algorithm on one particular problem: the object-background segmentation problem for video. It should be noted that the application of our algorithm is not limited to the above problem, the algorithm is generic and can be used to yield similar improvements in many other cases that involve dynamic change.

In this paper, we present a fast new fully dynamic algorithm for the st-mincut/max-flow problem. We show how this algorithm can be used to efficiently compute MAP estimates for dynamically changing MRF models of labeling problems in computer vision, such as image segmentation. Specifically, given the solution of the max-flow problem on a graph, we show how to efficiently compute the maximum flow in a modified version of the graph. Our experiments showed that the time taken by our algorithm is roughly proportional to the number of edges whose weights were different in the two graphs. We test the performance of our algorithm on one particular problem: the object-background segmentation problem for video and compare it with the best known st-mincut algorithm. The results show that the dynamic graph cut algorithm is much faster than its static counterpart and enables real time image segmentation. It should be noted that our method is generic and can be used to yield similar improvements in many other cases that involve dynamic change in the graph