Large Vision-Language Models (LVLMs) with "multimodal distractibility," where plausible but irrelevant visual or textual inputs cause significant drops in reasoning consistency and lead to unreliable outputs. This paper introduces a comprehensive framework to systematically diagnose, evaluate, and mitigate this critical challenge. We present three core components: the large-scale IR-VQA benchmark to surface these vulnerabilities across four paradigms; novel diagnostic metrics, Positive Consistency (PC) and Negative Consistency (NC), which move beyond standard accuracy to rigorously measure a model's reasoning stability; and the Relevance-Gated Multimodal Routing (RGMR) mechanism, a novel, lightweight module that proactively and dynamically filters distractions at inference time. Our experiments reveal that state-of-the-art models exhibit significant drops in consistency on IR-VQA. We demonstrate that finetuning on IR-VQA and deploying RGMR substantially improve model robustness where standard prompting fails. Our comprehensive analysis of model behaviors under different types of distractions and the underlying reasoning failures provides a clear path forward for developing more reliable multimodal systems.

Deep saliency models, which predict what parts of an image capture our attention, are often like black boxes. This limits their use, especially in areas where understanding why a model makes a decision is crucial. Our research tackles this challenge by developing an explainable saliency (XSal) model that not only identifies what is important in an image, but also explains its choices in a way that makes sense to humans. We achieve this by using vision-language models to reason about images and by focusing the model’s attention on the most crucial information using a contextual prioritization mechanism. Unlike prior approaches that rely on fixation descriptions or soft-attention based semantic aggregation, our method directly models the reasoning steps involved in saliency prediction, generating selectively prioritized explanations clarify why specific regions are prioritized. Comprehensive evaluations demonstrate the effectiveness of our model in generating high-quality saliency maps and coherent, contextually relevant explanations. This research is a step towards more transparent and trustworthy AI systems that can help us understand and navigate the world around us.

This paper introduces CameraCtrl II, a framework that enables large-scale dynamic scene exploration through a camera-controlled video diffusion model. Previous camera-conditioned video generative models suffer from diminished video dynamics and limited range of viewpoints when generating videos with large camera movement. We take an approach that progressively expands the generation of dynamic scenes -- first enhancing dynamic content within individual video clip, then extending this capability to create seamless explorations across broad viewpoint ranges. Specifically, we construct a dataset featuring a large degree of dynamics with camera parameter annotations for training while designing a lightweight camera injection module and training scheme to preserve dynamics of the pretrained models. Building on these improved single-clip techniques, we enable extended scene exploration by allowing users to iteratively specify camera trajectories for generating coherent video sequences. Experiments across diverse scenarios demonstrate that CameraCtrl Ii enables camera-controlled dynamic scene synthesis with substantially wider spatial exploration than previous approaches.

This paper presents a multimodal framework that attempts to unify visual understanding and generation within a shared discrete semantic representation. At its core is the Text-Aligned Tokenizer (TA-Tok), which converts images into discrete tokens using a text-aligned codebook projected from a large language model's (LLM) vocabulary. By integrating vision and text into a unified space with an expanded vocabulary, our multimodal LLM, **Tar**, enables cross-modal input and output through a shared interface, without the need for modality-specific designs. Additionally, we propose scale-adaptive encoding and decoding to balance efficiency and visual detail, along with a generative de-tokenizer to produce high-fidelity visual outputs. To address diverse decoding needs, we utilize two complementary de-tokenizers: a fast autoregressive model and a diffusion-based model. To enhance modality fusion, we investigate advanced pre-training tasks, demonstrating improvements in both visual understanding and generation. Experiments across benchmarks show that **Tar** matches or surpasses existing multimodal LLM methods, achieving faster convergence and greater training efficiency. All code, models, and data will be made publicly available.

A key challenge for machine intelligence is to learn new visual concepts without forgetting the previously acquired knowledge. Continual learning (CL) is aimed toward addressing this challenge. However, there still exists a gap between CL and human learning. In particular, humans are able to continually learn from the samples associated with known or unknown labels in their daily lives, whereas existing CL and semi-supervised CL (SSCL) methods assume that the training samples are associated with known labels. Specifically, we are interested in two questions: 1) how to utilize unrelated unlabeled data for the SSCL task and 2) how unlabeled data affect learning and catastrophic forgetting in the CL task. To explore these issues, we formulate a new SSCL method, which can be generically applied to existing CL models. Furthermore, we propose a novel gradient learner to learn from labeled data to predict gradients on unlabeled data. In this way, the unlabeled data can fit into the supervised CL framework. We extensively evaluate the proposed method on mainstream CL methods, adversarial CL (ACL), and semi-supervised learning (SSL) tasks. The proposed method achieves state-of-the-art performance on classification accuracy and backward transfer (BWT) in the CL setting while achieving the desired performance on classification accuracy in the SSL setting. This implies that the unlabeled images can enhance the generalizability of CL models on the predictive ability of unseen data and significantly alleviate catastrophic forgetting. The code is available at https://github.com/luoyan407/grad_prediction.git.

Understanding how attention varies across individuals has significant scientific and societal impacts. However, existing visual scanpath models treat attention uniformly, neglecting individual differences. To bridge this gap, this paper focuses on individualized scanpath prediction (ISP), a new attention modeling task that aims to accurately predict how different individuals shift their attention in diverse visual tasks. It proposes an ISP method featuring three novel technical components: (1) an observer encoder to characterize and integrate an observer's unique attention traits, (2) an observer-centric feature integration approach that holistically combines visual features, task guidance, and observer-specific characteristics, and (3) an adaptive fixation prioritization mechanism that refines scanpath predictions by dynamically prioritizing semantic feature maps based on individual observers' attention traits. These novel components allow scanpath models to effectively address the attention variations across different observers. Our method is generally applicable to different datasets, model architectures, and visual tasks, offering a comprehensive tool for transforming general scanpath models into individualized ones. Comprehensive evaluations using value-based and ranking-based metrics verify the method's effectiveness and generalizability.

Integrating information from vision and language modalities has sparked interesting applications in the fields of computer vision and natural language processing. Existing methods, though promising in tasks like image captioning and visual question answering, face challenges in understanding real-life issues and offering step-by-step solutions. In particular, they typically limit their scope to solutions with a sequential structure, thus ignoring complex inter-step dependencies. To bridge this gap, we propose a graph-based approach to vision-language problem solving. It leverages a novel integrated attention mechanism that jointly considers the importance of features within each step as well as across multiple steps. Together with a graph neural network method, this attention mechanism can be progressively learned to predict sequential and non-sequential solution graphs depending on the characterization of the problem-solving process. To tightly couple attention with the problem-solving procedure, we further design new learning objectives with attention metrics that quantify this integrated attention, which better aligns visual and language information within steps, and more accurately captures information flow between steps. Experimental results on VisualHow, a comprehensive dataset of varying solution structures, show significant improvements in predicting steps and dependencies, demonstrating the effectiveness of our approach in tackling various vision-language problems.

Humans have the innate capability to answer diverse questions, which is rooted in the natural ability to correlate different concepts based on their semantic relationships and decompose difficult problems into sub-tasks. On the contrary, existing visual reasoning methods assume training samples that capture every possible object and reasoning problem, and rely on black-boxed models that commonly exploit statistical priors. They have yet to develop the capability to address novel objects or spurious biases in real-world scenarios, and also fall short of interpreting the rationales behind their decisions. Inspired by humans' reasoning of the visual world, we tackle the aforementioned challenges from a compositional perspective, and propose an integral framework consisting of a principled object factorization method and a novel neural module network. Our factorization method decomposes objects based on their key characteristics, and automatically derives prototypes that represent a wide range of objects. With these prototypes encoding important semantics, the proposed network then correlates objects by measuring their similarity on a common semantic space and makes decisions with a compositional reasoning process. It is capable of answering questions with diverse objects regard-less of their availability during training, and overcoming the issues of biased question-answer distributions. In addition to the enhanced generalizability, our framework also provides an interpretable interface for understanding the decision-making process of models. Our code is available at https://github.com/szzexpoi/POEM.

Answering visual questions requires the ability to parse visual observations and correlate them with a variety of knowledge. Existing visual question answering (VQA) models either pay little attention to the role of knowledge or do not take into account the granularity of knowledge (e.g., attaching the color of "grassland" to "ground"). They have yet to develop the capability of modeling knowledge of multiple granularity, and are also vulnerable to spurious data biases. To fill the gap, this paper makes progresses from two distinct perspectives: (1) It presents a Hierarchical Concept Graph (HCG) that discriminates and associates multi-granularity concepts with a multi-layered hierarchical structure, aligning visual observations with knowledge across different levels to alleviate data biases. (2) To facilitate a comprehensive understanding of how knowledge contributes throughout the decision-making process, we further propose an interpretable Hierarchical Concept Neural Module Network (HCNMN). It explicitly propagates multi-granularity knowledge across the hierarchical structure and incorporates them with a sequence of reasoning steps, providing a transparent interface to elaborate on the integration of observations and knowledge. Through extensive experiments on multiple challenging datasets (i.e., GQA,VQA,FVQA,OK-VQA), we demonstrate the effectiveness of our method in answering questions in different scenarios. Our code is available at https://github.com/SuperJohnZhang/HCNMN.

This paper presents a fully integrated W-band multi-channel silicon-based radiometer array in a 65-nm CMOS process. Four channels are integrated to achieve higher resolution and sensitivity. A single channel consists of low noise amplifier (LNA), detector, and analog baseband (ABB), respectively. The four-stage LNA achieves a power gain of 29.9dB, a minimum noise figure (NF) of 5.0dB over the band of interest. The detector achieves responsivity of 2.34 kV/W and 23.9$\mathbf{pW}/\sqrt{\text{Hz}}$noise equivalent power (NEP). The radiometer achieves a responsivity of 42293 MV/W, an NEP of 0.21$\mathbf{fW}/\sqrt{\text{Hz}}$, and a temperature resolution (NETD) of 0.226 K with a 1 ms integration time. The total power consumption is 254.6 mW.

In recent years, deep saliency models have made significant progress in predicting human visual attention. However, the mechanisms behind their success remain largely unexplained due to the opaque nature of deep neural networks. In this paper, we present a novel analytic framework that sheds light on the implicit features learned by saliency models and provides principled interpretation and quantification of their contributions to saliency prediction. Our approach decomposes these implicit features into interpretable bases that are explicitly aligned with semantic attributes and reformulates saliency prediction as a weighted combination of probability maps connecting the bases and saliency. By applying our framework, we conduct extensive analyses from various perspectives, including the positive and negative weights of semantics, the impact of training data and architectural designs, the progressive influences of fine-tuning, and common error patterns of state-of-the-art deep saliency models. Additionally, we demonstrate the effectiveness of our framework by exploring visual attention characteristics in various application scenarios, such as the atypical attention of people with autism spectrum disorder, attention to emotion-eliciting stimuli, and attention evolution over time. Our code is publicly available at \url{https://github.com/szzexpoi/saliency_analysis}.

Attending selectively to emotion-eliciting stimuli is intrinsic to human vision. In this research, we investigate how emotion-elicitation features of images relate to human selective attention. We create the EMOtional attention dataset (EMOd). It is a set of diverse emotion-eliciting images, each with (1) eye-tracking data from 16 subjects, (2) image context labels at both object- and scene-level. Based on analyses of human perceptions of EMOd, we report an emotion prioritization effect: emotion-eliciting content draws stronger and earlier human attention than neutral content, but this advantage diminishes dramatically after initial fixation. We find that human attention is more focused on awe eliciting and aesthetic vehicle and animal scenes in EMOd. Aiming to model the above human attention behavior computationally, we design a deep neural network (CASNet II), which includes a channel weighting subnetwork that prioritizes emotion-eliciting objects, and an Atrous Spatial Pyramid Pooling (ASPP) structure that learns the relative importance of image regions at multiple scales. Visualizations and quantitative analyses demonstrate the model's ability to simulate human attention behavior, especially on emotion-eliciting content.

The learning process of deep learning methods usually updates the model’s parameters in multiple iterations. Each iteration can be viewed as the first-order approximation of Taylor’s series expansion. The remainder, which consists of higher-order terms, is usually ignored in the learning process for simplicity. This learning scheme empowers various multimedia-based applications, such as image retrieval, recommendation system, and video search. Generally, multimedia data (e.g.images) are semantics-rich and high-dimensional, hence the remainders of approximations are possibly non-zero. In this work, we consider that the remainder is informative and study how it affects the learning process. To this end, we propose a new learning approach, namely gradient adjustment learning (GAL), to leverage the knowledge learned from the past training iterations to adjust vanilla gradients, such that the remainders are minimized and the approximations are improved. The proposed GAL is model- and optimizer-agnostic, and is easy to adapt to the standard learning framework. It is evaluated on three tasks,i.e.image classification, object detection, and regression, with state-of-the-art models and optimizers. The experiments show that the proposed GAL consistently enhances the evaluated models, whereas the ablation studies validate various aspects of the proposed GAL. The code is available athttps://github.com/luoyan407/gradient_adjustment.git.

Effectiveness and interpretability are two essential properties for trustworthy AI systems. Most recent studies in visual reasoning are dedicated to improving the accuracy of predicted answers, and less attention is paid to explaining the rationales behind the decisions. As a result, they commonly take advantage of spurious biases instead of actually reasoning on the visual-textual data, and have yet developed the capability to explain their decision making by considering key information from both modalities. This paper aims to close the gap from three distinct perspectives: first, we define a new type of multi-modal explanations that explain the decisions by progressively traversing the reasoning process and grounding keywords in the images. We develop a functional program to sequentially ex-ecute different reasoning steps and construct a new dataset with 1,040,830 multi-modal explanations. Second, we iden-tify the critical need to tightly couple important components across the visual and textual modalities for explaining the decisions, and propose a novel explanation generation method that explicitly models the pairwise correspon-dence between words and regions of interest. It improves the visual grounding capability by a considerable margin, resulting in enhanced interpretability and reasoning performance. Finally, with our new data and method, we perform extensive analyses to study the effectiveness of our explanation under different settings, including multi-task learning and transfer learning. Our code and data are available at https://github.com/szzexpoi/rex.

Recent progress in the interdisciplinary studies of computer vision (CV) and natural language processing (NLP) has enabled the development of intelligent systems that can describe what they see and answer questions accordingly. However, despite showing usefulness in performing these vision-language tasks, existing methods still struggle in understanding real-life problems (i.e., how to do something) and suggesting step-by-step guidance to solve them. With an overarching goal of developing intelligent systems to assist humans in various daily activities, we propose VisualHow, a free-form and open-ended research that focuses on understanding a real-life problem and deriving its solution by incorporating key components across multiple modalities. We develop a new dataset with 20,028 real-life problems and 102,933 steps that constitute their solutions, where each step consists of both a visual illustration and a textual description that guide the problem solving. To establish better understanding of problems and solutions, we also provide annotations of multimodal attention that localizes important components across modalities and solution graphs that encapsulate different steps in structured representations. These data and annotations enable a family of new vision-language tasks that solve real-life problems. Through extensive experiments with representative models, we demonstrate their effectiveness on training and testing models for the new tasks, and there is significant scope for improvement by learning effective attention mechanisms. Our dataset and models are available at https://github.com/formidify/VisualHow.

Explainable visual question answering (VQA) models have been developed with neural modules and query-based knowledge incorporation to answer knowledge-requiring questions. Yet, most reasoning methods cannot effectively generate queries or incorporate external knowledge during the reasoning process, which may lead to suboptimal results. To bridge this research gap, we present Query and Attention Augmentation, a general approach that augments neural module networks to jointly reason about visual and external knowledge. To take both knowledge sources into account during reasoning, it parses the input question into a functional program with queries augmented through a novel reinforcement learning method, and jointly directs augmented attention to visual and external knowledge based on intermediate reasoning results. With extensive experiments on multiple VQA datasets, our method demonstrates significant performance, explainability, and generalizability over state-of-the-art models in answering questions requiring different extents of knowledge. Our source code is available at https://github.com/SuperJohnZhang/QAA.

While attention has been an increasingly popular component in deep neural networks to both interpret and boost the performance of models, little work has examined how attention progresses to accomplish a task and whether it is reasonable. In this work, we propose an Attention with Reasoning capability (AiR) framework that uses attention to understand and improve the process leading to task outcomes. We first define an evaluation metric based on a sequence of atomic reasoning operations, enabling a quantitative measurement of attention that considers the reasoning process. We then collect human eye-tracking and answer correctness data, and analyze various machine and human attention mechanisms on their reasoning capability and how they impact task performance. To improve the attention and reasoning ability of visual question answering models, we propose to supervise the learning of attention progressively along the reasoning process and to differentiate the correct and incorrect attention patterns. We demonstrate the effectiveness of the proposed framework in analyzing and modeling attention with better reasoning capability and task performance. The code and data are available at https://github.com/szzexpoi/AiR.

Attention has been an important mechanism for both humans and computer vision systems. While state-of-the-art models to predict attention focus on estimating a static probabilistic saliency map with free-viewing behavior, real-life scenarios are filled with tasks of varying types and complexities, and visual exploration is a temporal process that contributes to task performance. To bridge the gap, we conduct a first study to understand and predict the temporal sequences of eye fixations (a.k.a. scanpaths) during performing general tasks, and examine how scanpaths affect task performance. We present a new deep reinforcement learning method to predict scanpaths leading to different performances in visual question answering. Conditioned on a task guidance map, the proposed model learns question-specific attention patterns to generate scanpaths. It addresses the exposure bias in scanpath prediction with self-critical sequence training and designs a Consistency-Divergence loss to generate distinguishable scanpaths between correct and incorrect answers. The proposed model not only accurately predicts the spatio-temporal patterns of human behavior in visual question answering, such as fixation position, duration, and order, but also generalizes to free-viewing and visual search tasks, achieving human-level performance in all tasks and significantly outperforming the state of the art.

Existing explainable and explicit visual reasoning methods only perform reasoning based on visual evidence but do not take into account knowledge beyond what is in the visual scene. To addresses the knowledge gap between visual reasoning methods and the semantic complexity of real-world images, we present the first explicit visual reasoning method that incorporates external knowledge and models high-order relational attention for improved generalizability and explainability. Specifically, we propose a knowledge incorporation network that explicitly creates and includes new graph nodes for entities and predicates from external knowledge bases to enrich the semantics of the scene graph used in explicit reasoning. We then create a novel Graph-Relate module to perform high-order relational attention on the enriched scene graph. By explicitly introducing structured external knowledge and high-order relational attention, our method demonstrates significant generalizability and explainability over the state-of-the-art visual reasoning approaches on the GQA and VQAv2 datasets.

Image captioning models depend on training with paired image-text corpora, which poses various challenges in describing images containing novel objects absent from the training data. While previous novel object captioning methods rely on external image taggers or object detectors to describe novel objects, we present the Attention-based Novel Object Captioner (ANOC) that complements novel object captioners with human attention features that characterize generally important information independent of tasks. It introduces a gating mechanism that adaptively incorporates human attention with self-learned machine attention, with a Constrained Self-Critical Sequence Training method to address the exposure bias while maintaining constraints of novel object descriptions. Extensive experiments conducted on the nocaps and Held-Out COCO datasets demonstrate that our method considerably outperforms the state-of-the-art novel object captioners. Our source code is available at https://github.com/chenxy99/ANOC.

In this workshop, we are addressing the trustworthy AI issues for Multimedia Computing. We aim to bring together researchers in the trustworthy aspects of Multimedia Computing and facilitate discussions in injecting trusts into multimedia to develop trustworthy AI techniques that are reliable and acceptable to multimedia researchers and practitioners. Our scope is at the conjunction of multimedia, computer vision and trustworthy AI, including Explainability, Robustness and Safety, Data Privacy, Accountability and Transparency, and Fairness.

Understanding the trustworthiness of a prediction yielded by a classifier is critical for the safe and effective use of AI models. Prior efforts have been proven to be reliable on small-scale datasets. In this work, we study the problem of predicting trustworthiness on real-world large-scale datasets, where the task is more challenging due to high-dimensional features, diverse visual concepts, and a large number of samples. In such a setting, we observe that the trustworthiness predictors trained with prior-art loss functions, i.e., the cross entropy loss, focal loss, and true class probability confidence loss, are prone to view both correct predictions and incorrect predictions to be trustworthy. The reasons are two-fold. Firstly, correct predictions are generally dominant over incorrect predictions. Secondly, due to the data complexity, it is challenging to differentiate the incorrect predictions from the correct ones on real-world large-scale datasets. To improve the generalizability of trustworthiness predictors, we propose a novel steep slope loss to separate the features w.r.t. correct predictions from the ones w.r.t. incorrect predictions by two slide-like curves that oppose each other. The proposed loss is evaluated with two representative deep learning models, i.e., Vision Transformer and ResNet, as trustworthiness predictors. We conduct comprehensive experiments and analyses on ImageNet, which show that the proposed loss effectively improves the generalizability of trustworthiness predictors. The code and pre-trained trustworthiness predictors for reproducibility are available at \url{https://github.com/luoyan407/predict_trustworthiness}.

The development of large-scale image-captioning datasets is expensive, while the abundance of unpaired images and text corpus can potentially help reduce the efforts of manual annotation. In this paper, we study the few-shot image captioning problem that only requires a small amount of annotated image-caption pairs. We propose an ensemble- based self-distillation method that allows image captioning models to be trained with unpaired images and captions. The ensemble consists of multiple base models trained with different data samples in each iteration. For learning from unpaired images, we generate multiple pseudo captions with the ensemble and allocate different weights according to their confidence levels. For learning from unpaired captions, we propose a simple yet effective pseudo feature generation method based on Gradient Descent. The pseudo captions and pseudo features from the ensemble are used to train the base models in future iterations. The proposed method is general over different image captioning models and datasets. Our experiments demonstrate significant performance improvements and meaningful captions generated with only 1% of paired training data. Source code is available at https://github.com/chenxy99/SD-FSIC.

The last decades have seen great progress in saliency prediction, with the success of deep neural networks that are able to encode high-level semantics. Yet, while humans have the innate capability in leveraging their knowledge to decide where to look (e.g. people pay more attention to familiar faces such as celebrities), saliency prediction models have only been trained with large eye-tracking datasets. This work proposes to bridge this gap by explicitly incorporating external knowledge for saliency models as humans do. We develop networks that learn to highlight regions by incorporating prior knowledge of semantic relationships, be it general or domain-specific, depending on the task of interest. At the core of the method is a new Graph Semantic Saliency Network (GraSSNet) that constructs a graph that encodes semantic relationships learned from external knowledge. A Spatial Graph Attention Network is then developed to update saliency features based on the learned graph. Experiments show that the proposed model learns to predict saliency from the external knowledge and outperforms the state-of-the-art on four saliency benchmarks.

One of the well-known challenges in computer vision tasks is the visual diversity of images, which could result in an agreement or disagreement between the learned knowledge and the visual content exhibited by the current observation. In this work, we first define such an agreement in a concepts learning process as congruency. Formally, given a particular task and sufficiently large dataset, the congruency issue occurs in the learning process whereby the task-specific semantics in the training data are highly varying. We propose a Direction Concentration Learning (DCL) method to improve congruency in the learning process, where enhancing congruency influences the convergence path to be less circuitous. The experimental results show that the proposed DCL method generalizes to state-of-the-art models and optimizers, as well as improves the performances of saliency prediction task, continual learning task, and classification task. Moreover, it helps mitigate the catastrophic forgetting problem in the continual learning task. The code is publicly available at https://github.com/luoyan407/congruency.

While most visual attention studies focus on bottom-up attention with restricted field-of-view, real-life situations are filled with embodied vision tasks. The role of attention is more significant in the latter due to the information overload, and attention to the most important regions is critical to the success of tasks. The effects of visual attention on task performance in this context have also been widely ignored. This research addresses a number of challenges to bridge this research gap, on both the data and model aspects. Specifically, we introduce the first dataset of top-down attention in immersive scenes. The Immersive Question-directed Visual Attention (IQVA) dataset features visual attention and corresponding task performance (i.e., answer correctness). It consists of 975 questions and answers collected from people viewing 360° videos in a head-mounted display. Analyses of the data demonstrate a significant correlation between people's task performance and their eye movements, suggesting the role of attention in task performance. With that, a neural network is developed to encode the differences of correct and incorrect attention and jointly predict the two. The proposed attention model for the first time takes into account answer correctness, whose outputs naturally distinguish important regions from distractions. This study with new data and features may enable new tasks that leverage attention and answer correctness, and inspire new research that reveals the process behind decision making in performing various tasks.

The computer vision community is witnessing an unprecedented rate of new tasks being proposed and addressed, thanks to the deep convolutional networks' capability to find complex mappings from X to Y. The advent of each task often accompanies the release of a large-scale annotated dataset, for supervised training of deep network. However, it is expensive and time-consuming to manually label sufficient amount of training data. Therefore, it is important to develop algorithms that can leverage off-the-shelf labeled dataset to learn useful knowledge for the target task. While previous works mostly focus on transfer learning from a single source, we study multi-source transfer across domains and tasks (MS-DTT), in a semi-supervised setting. We propose GradMix, a model-agnostic method applicable to any model trained with gradient-based learning rule, to transfer knowledge via gradient descent by weighting and mixing the gradients from all sources during training. GradMix follows a meta-learning objective, which assigns layer-wise weights to the source gradients, such that the combined gradient follows the direction that minimize the loss for a small set of samples from the target dataset. In addition, we propose to adaptively adjust the learning rate for each mini-batch based on its importance to the target task, and a pseudo-labeling method to leverage the unlabeled samples in the target domain. We conduct MS-DTT experiments on two tasks: digit recognition and action recognition, and demonstrate the advantageous performance of the proposed method against multiple baselines.

Bridging vision and natural language is a longstanding goal in computer vision and multimedia research. While earlier works focus on generating a single-sentence description for visual content, recent works have studied paragraph generation. In this paper, we introduce the problem of video storytelling, which aims at generating coherent and succinct stories for long videos. Video storytelling introduces new challenges, mainly due to the diversity of the story and the length and complexity of the video. We propose novel methods to address the challenges. First, we propose a context-aware framework for multimodal embedding learning, where we design a residual bidirectional recurrent neural network to leverage contextual information from past and future. The multimodal embedding is then used to retrieve sentences for video clips. Second, we propose a Narrator model to select clips that are representative of the underlying storyline. The Narrator is formulated as a reinforcement learning agent, which is trained by directly optimizing the textual metric of the generated story. We evaluate our method on the video story dataset, a new dataset that we have collected to enable the study. We compare our method with multiple state-of-the-art baselines and show that our method achieves better performance, in terms of quantitative measures and user study.

The recent advances in instance-level detection tasks lay strong foundation for genuine comprehension of the visual scenes. However, the ability to fully comprehend a social scene is still in its preliminary stage. In this work, we focus on detecting human-object interactions (HOIs) in social scene images, which is demanding in terms of research and increasingly useful for practical applications. To undertake social tasks interacting with objects, humans direct their attention and move their body based on their intention. Based on this observation, we provide a unique computational perspective to explore human intention in HOI detection. Specifically, the proposed human intention-driven HOI detection (iHOI) framework models human pose with the relative distances from body joints to the object instances. It also utilizes human gaze to guide the attended contextual regions in a weakly-supervised setting. In addition, we propose a hard negative sampling strategy to address the problem of mis-grouping. We perform extensive experiments on two benchmark datasets, namely V-COCO and HICO-DET. The efficacy of each proposed component has also been validated.

Intraclass compactness and interclass separability are crucial indicators to measure the effectiveness of a model to produce discriminative features, where intraclass compactness indicates how close the features with the same label are to each other and interclass separability indicates how far away the features with different labels are. In this paper, we investigate intraclass compactness and interclass separability of features learned by convolutional networks and propose a Gaussian-based softmax ( G -softmax) function that can effectively improve intraclass compactness and interclass separability. The proposed function is simple to implement and can easily replace the softmax function. We evaluate the proposed G -softmax function on classification data sets (i.e., CIFAR-10, CIFAR-100, and Tiny ImageNet) and on multilabel classification data sets (i.e., MS COCO and NUS-WIDE). The experimental results show that the proposed G -softmax function improves the state-of-the-art models across all evaluated data sets. In addition, the analysis of the intraclass compactness and interclass separability demonstrates the advantages of the proposed function over the softmax function, which is consistent with the performance improvement. More importantly, we observe that high intraclass compactness and interclass separability are linearly correlated with average precision on MS COCO and NUS-WIDE. This implies that the improvement of intraclass compactness and interclass separability would lead to the improvement of average precision.

Despite the success of deep neural networks (DNNs) in image classification tasks, the human-level performance relies on massive training data with high-quality manual annotations, which are expensive and time-consuming to collect. There exist many inexpensive data sources on the web, but they tend to contain inaccurate labels. Training on noisy labeled datasets causes performance degradation because DNNs can easily overfit to the label noise. To overcome this problem, we propose a noise-tolerant training algorithm, where a meta-learning update is performed prior to conventional gradient update. The proposed meta-learning method simulates actual training by generating synthetic noisy labels, and train the model such that after one gradient update using each set of synthetic noisy labels, the model does not overfit to the specific noise. We conduct extensive experiments on the noisy CIFAR-10 dataset and the Clothing1M dataset. The results demonstrate the advantageous performance of the proposed method compared to several state-of-the-art baselines.

The recent advances in instance-level detection tasks lay a strong foundation for automated visual scenes understanding. However, the ability to fully comprehend a social scene still eludes us. In this work, we focus on detecting human-object interactions (HOIs) in images, an essential step towards deeper scene understanding. HOI detection aims to localize human and objects, as well as to identify the complex interactions between them. Innate in practical problems with large label space, HOI categories exhibit a long-tail distribution, i.e., there exist some rare categories with very few training samples. Given the key observation that HOIs contain intrinsic semantic regularities despite they are visually diverse, we tackle the challenge of long-tail HOI categories by modeling the underlying regularities among verbs and objects in HOIs as well as general relationships. In particular, we construct a knowledge graph based on the ground-truth annotations of training dataset and external source. In contrast to direct knowledge incorporation, we address the necessity of dynamic image-specific knowledge retrieval by multi-modal learning, which leads to an enhanced semantic embedding space for HOI comprehension. The proposed method shows improved performance on V-COCO and HICO-DET benchmarks, especially when predicting the rare HOI categories.

This paper presents a novel framework for automatic and quantitative screening of autism spectrum disorder (ASD). It is motivated to address two issues in the current clinical settings: 1) short of clinical resources with the prevalence of ASD (1.7% in the United States), and 2) subjectivity of ASD screening. This work differentiates itself with three unique features: first, it proposes an ASD screening with privileged modality framework that integrates information from two behavioral modalities during training and improves the performance on each single modality at testing. The proposed framework does not require overlap in subjects between the modalities. Second, it develops the first computational model to classify people with ASD using a photo-taking task where subjects freely explore their environment in a more ecological setting. Photo-taking reveals attentional preference of subjects, differentiating people with ASD from healthy people, and is also easy to implement in real-world clinical settings without requiring advanced diagnostic instruments. Third, this study for the first time takes advantage of the temporal information in eye movements while viewing images, encoding more detailed behavioral differences between ASD people and healthy controls. Experiments show that our ASD screening models can achieve superior performance, outperforming the previous state-of-the-art methods by a considerable margin. Moreover, our framework using diverse modalities demonstrates performance improvement on both the photo-taking and image-viewing tasks, providing a general paradigm that takes in multiple sources of behavioral data for a more accurate ASD screening. The framework is also applicable to various scenarios where one-to-one pairwise relationship is difficult to obtain across different modalities.

Redundancy is a fundamental characteristic of many biological processes such as those in the genetic, visual, muscular, and nervous systems, yet its driven mechanism has not been fully comprehended. Until recently, the only understanding of redundancy is as a mean to attain fault tolerance, which is reflected in the design of many man-made systems. On the contrary, our previous work on redundant sensing (RS) has demonstrated an example where redundancy can be engineered solely for enhancing accuracy and precision. The design was inspired by the binocular structure of human vision, which we believe may share a similar operation. In this letter, we present a unified theory describing how such utilization of redundancy is feasible through two complementary mechanisms: representational redundancy (RPR) and entangled redundancy (ETR). We also point out two additional examples where our new understanding of redundancy can be applied to justify a system's superior performance. One is the human musculoskeletal system (HMS), a biological instance, and the other is the deep residual neural network (ResNet), an artificial counterpart. We envision that our theory would provide a framework for the future development of bio-inspired redundant artificial systems, as well as assist studies of the fundamental mechanisms governing various biological processes.

Recent surge of Convolutional Neural Networks (CNNs) has brought successes among various applications. However, these successes are accompanied by a significant increase in computational cost and the demand for computational resources, which critically hampers the utilization of complex CNNs on devices with limited computational power. In this work, we propose a feature representation based layer-wise pruning method that aims at reducing complex CNNs to more compact ones with equivalent performance. Different from previous parameter pruning methods that conduct connection-wise or filter-wise pruning based on weight information, our method determines redundant parameters by investigating the features learned in the convolutional layers and the pruning process is operated at a layer level. Experiments demonstrate that the proposed method is able to significantly reduce computational cost and the pruned models achieve equivalent or even better performance compared to the original models on various datasets.

We introduce a new problem of gaze anticipation on future frames which extends the conventional gaze prediction problem to go beyond current frames. To solve this problem, we propose a new generative adversarial network based model, Deep Future Gaze (DFG), encompassing two pathways: DFG-P is to anticipate gaze prior maps conditioned on the input frame which provides task influences; DFG-G is to learn to model both semantic and motion information in future frame generation. DFG-P and DFG-G are then fused to anticipate future gazes. DFG-G consists of two networks: a generator and a discriminator. The generator uses a two-stream spatial-temporal convolution architecture (3D-CNN) for explicitly untangling the foreground and background to generate future frames. It then attaches another 3D-CNN for gaze anticipation based on these synthetic frames. The discriminator plays against the generator by distinguishing the synthetic frames of the generator from the real frames. Experimental results on the publicly available egocentric and third person video datasets show that DFG significantly outperforms all competitive baselines. We also demonstrate that DFG achieves better performance of gaze prediction on current frames in egocentric and third person videos than state-of-the-art methods.

When looking at an image, humans shift their attention toward interesting regions, making sequences of eye fixations. When describing an image, they also come up with simple sentences that highlight the key elements in the scene. What is the correlation between where people look and what they describe in an image? To investigate this problem intuitively, we develop a visual analytics system, CapVis, to look into visual attention and image captioning, two types of subjective annotations that are relatively task-free and natural. Using these annotations, we propose a word-weighting scheme to extract visual and verbal saliency ranks to compare against each other. In our approach, a number of low-level and semantic-level features relevant to visual-verbal saliency consistency are proposed and visualized for a better understanding of image content. Our method also shows the different ways that a human and a computational model look at and describe images, which provides reliable information for a captioning model. Experiment also shows that the visualized feature can be integrated into a computational model to effectively predict the consistency between the two modalities on an image dataset with both types of annotations.

Image sentiment influences visual perception. Emotion-eliciting stimuli such as happy faces and poisonous snakes are generally prioritized in human attention. However, little research has evaluated the interrelationships of image sentiment and visual saliency. In this paper, we present the first study to focus on the relation between emotional properties of an image and visual attention. We first create the EMOtional attention dataset (EMOd). It is a diverse set of emotion-eliciting images, and each image has (1) eye-tracking data collected from 16 subjects, (2) intensive image context labels including object contour, object sentiment, object semantic category, and high-level perceptual attributes such as image aesthetics and elicited emotions. We perform extensive analyses on EMOd to identify how image sentiment relates to human attention. We discover an emotion prioritization effect: for our images, emotion-eliciting content attracts human attention strongly, but such advantage diminishes dramatically after initial fixation. Aiming to model the human emotion prioritization computationally, we design a deep neural network for saliency prediction, which includes a novel subnetwork that learns the spatial and semantic context of the image scene. The proposed network outperforms the state-of-the-art on three benchmark datasets, by effectively capturing the relative importance of human attention within an image. The code, models, and dataset are available online at https://nus-sesame.top/emotionalattention/.

Egocentric spatial memory (ESM) defines a memory system with encoding, storing, recognizing and recalling the spatial information about the environment from an egocentric perspective. We introduce an integrated deep neural network architecture for modeling ESM. It learns to estimate the occupancy state of the world and progressively construct top-down 2D global maps from egocentric views in a spatially extended environment. During the exploration, our proposed ESM model updates belief of the global map based on local observations using a recurrent neural network. It also augments the local mapping with a novel external memory to encode and store latent representations of the visited places over longterm exploration in large environments which enables agents to perform place recognition and hence, loop closure. Our proposed ESM network contributes in the following aspects: (1) without feature engineering, our model predicts free space based on egocentric views efficiently in an end-to-end manner; (2) different from other deep learning-based mapping system, ESMN deals with continuous actions and states which is vitally important for robotic control in real applications. In the experiments, we demonstrate its accurate and robust global mapping capacities in 3D virtual mazes and realistic indoor environments by comparing with several competitive baselines.

The recent success in human action recognition with deep learning methods mostly adopt the supervised learning paradigm, which requires significant amount of manually labeled data to achieve good performance. However, label collection is an expensive and time-consuming process. In this work, we propose an unsupervised learning framework, which exploits unlabeled data to learn video representations. Different from previous works in video representation learning, our unsupervised learning task is to predict 3D motion in multiple target views using video representation from a source view. By learning to extrapolate cross-view motions, the representation can capture view-invariant motion dynamics which is discriminative for the action. In addition, we propose a view-adversarial training method to enhance learning of view-invariant features. We demonstrate the effectiveness of the learned representations for action recognition on multiple datasets.

Visual realism is defined as the extent to which an image appears to people as a photo rather than computer generated. Assessing visual realism is important in applications like computer graphics rendering and photo retouching. However, current realism evaluation approaches use either labor-intensive human judgments or automated algorithms largely dependent on comparing renderings to reference images. We develop a reference-free computational framework for visual realism prediction to overcome these constraints. First, we construct a benchmark dataset of 2,520 images with comprehensive human annotated attributes. From statistical modeling on this data, we identify image attributes most relevant for visual realism. We propose both empirically-based (guided by our statistical modeling of human data) and deep convolutional neural network models to predict visual realism of images. Our framework has the following advantages: (1) it creates an interpretable and concise empirical model that characterizes human perception of visual realism; (2) it links computational features to latent factors of human image perception.

We introduce a new problem of gaze anticipation on egocentric videos. This substantially extends the conventional gaze prediction problem to future frames by no longer confining it on the current frame. To solve this problem, we propose a new generative adversarial neural network based model, Deep Future Gaze (DFG). DFG generates multiple future frames conditioned on the single current frame and anticipates corresponding future gazes in next few seconds. It consists of two networks: generator and discriminator. The generator uses a two-stream spatial temporal convolution architecture (3D-CNN) explicitly untangling the foreground and the background to generate future frames. It then attaches another 3D-CNN for gaze anticipation based on these synthetic frames. The discriminator plays against the generator by differentiating the synthetic frames of the generator from the real frames. Through competition with discriminator, the generator progressively improves quality of the future frames and thus anticipates future gaze better. Experimental results on the publicly available egocentric datasets show that DFG significantly outperforms all well-established baselines. Moreover, we demonstrate that DFG achieves better performance of gaze prediction on current frames than state-of-the-art methods. This is due to benefiting from learning motion discriminative representations in frame generation. We further contribute a new egocentric dataset (OST) in the object search task. DFG also achieves the best performance for this challenging dataset.

This paper presents a novel method for quantitative and objective diagnoses of Autism Spectrum Disorder (ASD) using eye tracking and deep neural networks. ASD is prevalent, with 1.5% of people in the US. The lack of clinical resources for early diagnoses has been a long-lasting issue. This work differentiates itself with three unique features: first, the proposed approach is data-driven and free of assumptions, important for new discoveries in understanding ASD as well as other neurodevelopmental disorders. Second, we concentrate our analyses on the differences in eye movement patterns between healthy people and those with ASD. An image selection method based on Fisher scores allows feature learning with the most discriminative contents, leading to efficient and accurate diagnoses. Third, we leverage the recent advances in deep neural networks for both prediction and visualization. Experimental results show the superior performance of our method in terms of multiple evaluation metrics used in diagnostic tests.

Since the beginning of early civilizations, social relationships derived from each individual fundamentally form the basis of social structure in our daily life. In the computer vision literature, much progress has been made in scene understanding, such as object detection and scene parsing. Recent research focuses on the relationship between objects based on its functionality and geometrical relations. In this work, we aim to study the problem of social relationship recognition, in still images. We have proposed a dual-glance model for social relationship recognition, where the first glance fixates at the individual pair of interest and the second glance deploys attention mechanism to explore contextual cues. We have also collected a new large scale People in Social Context (PISC) dataset, which comprises of 22,670 images and 76,568 annotated samples from 9 types of social relationship. We provide benchmark results on the PISC dataset, and qualitatively demonstrate the efficacy of the proposed model.

Top-down attention plays an important role in guidance of human attention in real-world scenarios, but less efforts in computational modeing of visual attention has been put on it. Inspired by the mechanisms of top-down attention in human visual perception, we propose a multi-layer linear model of top-down attention to modulate bottom-up saliency maps actively. The first layer is a linear regression model which combines the bottom-up saliency maps on various visual features and objects. A contextual dependent upper layer is introduced to tune the parameters of the lower layer model adaptively. Finally, a mask of selection history is applied to the fused attention map to bias the attention selection towards the task related regions. Efficient learning algorithm with single-pass polynomial complexity is derived. We evaluate our model on a set of natural egocentric videos captured from a wearable glass in real-world environments. Our model outperforms the baseline and state-of-the-art bottom-up saliency models.

A novel deep convolution neural network is proposed to predict gaze on current frames in egocentric videos. Inspired by human visual system, we introduce a fovea module responsible for sharp central vision and name our model as Foveated Neural Network (FNN). The retina-like visual inputs from the region of interest on the previous frame are analysed and encoded. The fusion of the hidden representations of the previous frame and the feature maps of the current frame guides the gaze prediction on the current frame. In order to simulate motion, we also include the dense optical flow between these adjacent frames as additional input. Experimental results show that FNN outperforms the state-of-the-art algorithms in the publicly available egocentric dataset. The analysis of FNN demonstrates that the hidden representations of the foveated visual input from the previous frame as well as the motion information between adjacent frames are efficient in improving gaze prediction performance in egocentric videos.

Automated assessment of visual sentiment has many applications, such as monitoring social media and facilitating online advertising. In current research on automated visual sentiment assessment, images are mainly input and processed as a whole. However, human attention is biased, and a focal region with high acuity can disproportionately influence visual sentiment. To investigate how attention influences visual sentiment, we conducted experiments that reveal critical insights into human perception. We discover that negative sentiments are elicited by the focal region without a notable influence of contextual information, whereas positive sentiments are influenced by both focal and contextual information. Building on these insights, we create new deep convolutional neural networks for sentiment prediction that have additional channels devoted to encoding focal information. On two benchmark datasets, the proposed models demonstrate superior performance compared with the state-of-the-art methods. Extensive visualizations and statistical analyses indicate that the focal channels are more effective on images with focal objects, especially for images that also elicit negative sentiments.

Training deep learning based video classifiers for action recognition requires a large amount of labeled videos. The labeling process is labor-intensive and time-consuming. On the other hand, large amount of weakly-labeled images are uploaded to the Internet by users everyday. To harness the rich and highly diverse set of Web images, a scalable approach is to crawl these images to train deep learning based classifier, such as Convolutional Neural Networks (CNN). However, due to the domain shift problem, the performance of Web images trained deep classifiers tend to degrade when directly deployed to videos. One way to address this problem is to fine-tune the trained models on videos, but sufficient amount of annotated videos are still required. In this work, we propose a novel approach to transfer knowledge from image domain to video domain. The proposed method can adapt to the target domain (i.e. video data) with limited amount of training data. Our method maps the video frames into a low-dimensional feature space using the class-discriminative spatial attention map for CNNs. We design a novel Siamese EnergyNet structure to learn energy functions on the attention maps by jointly optimizing two loss functions, such that the attention map corresponding to a ground truth concept would have higher energy. We conduct extensive experiments on two challenging video recognition datasets (i.e. TVHI and UCF101), and demonstrate the efficacy of our proposed method.

Previous works have suggested the role of scene information in directing gaze. The structure of a scene provides global contextual information that complements local object information in saliency prediction. In this study, we explore how scene envelopes such as openness, depth, and perspective affect visual attention in natural outdoor images. To facilitate this study, an eye tracking dataset is first built with 500 natural scene images and eye tracking data with 15 subjects free-viewing the images. We make observations on scene layout properties and propose a set of scene structural features relating to visual attention. We further integrate features from deep neural networks and use the set of complementary features for saliency prediction. Our features are independent of and can work together with many computational modules, and this work demonstrates the use of Multiple kernel learning (MKL) as an example to integrate the features at low- and high-levels. Experimental results demonstrate that our model outperforms existing methods and our scene structural features can improve the performance of other saliency models in outdoor scenes.

When looking at an image, humans shift their attention towards interesting regions, making sequences of eye fixations. When describing an image, they also come up with simple sentences that highlight the key elements in the scene. What is the correlation between where people look and what they describe in an image? To investigate this problem, we look into eye fixations and image captions, two types of subjective annotations that are relatively task-free and natural. From the annotations, we extract visual and verbal saliency ranks to compare against each other. We then propose a number of low-level and semantic-level features relevant to the visualverbal consistency. Integrated into a computational model, the proposed features effectively predict the consistency between the two modalities on a large dataset with both types of annotations, namely SALICON [1].

Our goal is to automatically detect which direction a child is facing based on a single, simple overhead picture, and track that direction across time. Engaging in joint attention, which is the shared focus of two individuals on some object of interest, is a strong cue of typically developing children, and the lack thereof can be an indicator of autism spectrum disorder or other pervasive developmental disorder. Therefore, the goal of many psychology experiments with children is to determine when, for how long, and towards what the child looks after some bid for attention or reaction. While much research looks for the orientation of faces based on frontal or profile pictures, or non-morphable, larger objects like cars, fewer studies work in the setting of minimally-invasive overhead person gaze or orientation detection. To automatically detect the child's orientation during a human-robot interaction experiment, we mount a camera on the ceiling of a child development laboratory and analyze the video footage. We use multiple kernel learning on eight potential orientation directions to determine a child's orientation during the video recorded interaction. We also contribute the labelled dataset we used on this challenging problem.

In many sub-fields, researchers collect datasets of human ground truth that are used to create a new algorithm. For example, in research on image perception, datasets have been collected for topics such as what makes an image aesthetic or memorable. Despite high costs for human data collection, datasets are infrequently reused beyond their own fields of interest. Moreover, the algorithms built from them are domain-specific (predict a small set of attributes) and usually unconnected to one another. In this paper, we present a paradigm for building generalized and expandable models of human image perception. First, we fuse multiple fragmented and partially-overlapping datasets through data imputation. We then create a theoretically-structured statistical model of human image perception that is fit to the fused datasets. The resulting model has many advantages. (1) It is generalized, going beyond the content of the constituent datasets, and can be easily expanded by fusing additional datasets. (2) It provides a new ontology usable as a network to expand human data in a cost-effective way. (3) It can guide the design of a generalized computational algorithm for multi-dimensional visual perception. Indeed, experimental results show that a model-based algorithm outperforms state-of-the-art methods on predicting visual sentiment, visual realism and interestingness. Our paradigm can be used in various visual tasks (e.g., video summarization).

Most state-of-the-art visual attention models estimate the probability distribution of fixating the eyes in a location of the image, the so-called saliency maps. Yet, these models do not predict the temporal sequence of eye fixations, which may be valuable for better predicting the human eye fixations, as well as for understanding the role of the different cues during visual exploration. In this paper, we present a method for predicting the sequence of human eye fixations, which is learned from the recorded human eye-tracking data. We use least-squares policy iteration (LSPI) to learn a visual exploration policy that mimics the recorded eye-fixation examples. The model uses a different set of parameters for the different stages of visual exploration that capture the importance of the cues during the scanpath. In a series of experiments, we demonstrate the effectiveness of using LSPI for combining multiple cues at different stages of the scanpath. The learned parameters suggest that the low-level and high-level cues (semantics) are similarly important at the first eye fixation of the scanpath, and the contribution of high-level cues keeps increasing during the visual exploration. Results show that our approach obtains the state-of-the-art performances on two challenging data sets: 1) OSIE data set and 2) MIT data set.

Ming Jiang1 [email protected] Xavier Boix1,3 [email protected] Juan Xu1 [email protected] Gemma Roig2,3 [email protected] Luc Van Gool3 [email protected] Qi Zhao1 [email protected] 1 Department of Electrical and Computer Engineering National University of Singapore Singapore 2 CBMM, LCSL Massachusetts Institute of Technology Istituto Italiano di Tecnologia Cambridge, MA 3 Computer Vision Laboratory ETH Zurich Switzerland

Saliency in Context (SALICON) is an ongoing effort that aims at understanding and predicting visual attention. This paper presents a new method to collect large-scale human data during natural explorations on images. While current datasets present a rich set of images and task-specific annotations such as category labels and object segments, this work focuses on recording and logging how humans shift their attention during visual exploration. The goal is to offer new possibilities to (1) complement task-specific annotations to advance the ultimate goal in visual understanding, and (2) understand visual attention and learn saliency models, all with human attentional data at a much larger scale. We designed a mouse-contingent multi-resolutional paradigm based on neurophysiological and psychophysical studies of peripheral vision, to simulate the natural viewing behavior of humans. The new paradigm allowed using a general-purpose mouse instead of an eye tracker to record viewing behaviors, thus enabling large-scale data collection. The paradigm was validated with controlled laboratory as well as large-scale online data. We report in this paper a proof-of-concept SALICON dataset of human “free-viewing” data on 10,000 images from the Microsoft COCO (MS COCO) dataset with rich contextual information. We evaluated the use of the collected data in the context of saliency prediction, and demonstrated them a good source as ground truth for the evaluation of saliency algorithms.

Recently, an increasing number of works have proposed to learn visual saliency by leveraging human fixations. However, the collection of human fixations is time consuming and the existing eye tracking datasets are generally small when compared with other domains. Thus, it contains a certain degree of dataset bias due to the large image variations (e.g., outdoor scenes vs. emotion-evoking images). In the learning based saliency prediction literature, most models are trained and evaluated within the same dataset and cross dataset validation is not yet a common practice. Instead of directly applying model learned from another dataset in cross dataset fashion, it is better to transfer the prior knowledge obtained from one dataset to improve the training and prediction on another. In addition, since new datasets are built and shared in the community from time to time, it would be good not to retrain the entire model when new data are added. To address these problems, we proposed a new learning based saliency model, namely Label Consistent Quadratic Surrogate algorithm, which employs an iterative online algorithm to learn a sparse dictionary with label consistent constraint. The advantages of the proposed model are three-folds: (1) the quadratic surrogate function guarantees convergence at each iteration, (2) the label consistent constraint enforces the predicted sparse code to be discriminative, and (3) the online properties enable the proposed algorithm to adapt existing model with new data without retraining. As shown in this work, the proposed saliency model achieves better performance than the state-of-the-art saliency models.

Saliency in Context (SALICON) is an ongoing effort that aims at understanding and predicting visual attention. Conventional saliency models typically rely on low-level image statistics to predict human fixations. While these models perform significantly better than chance, there is still a large gap between model prediction and human behavior. This gap is largely due to the limited capability of models in predicting eye fixations with strong semantic content, the so-called semantic gap. This paper presents a focused study to narrow the semantic gap with an architecture based on Deep Neural Network (DNN). It leverages the representational power of high-level semantics encoded in DNNs pretrained for object recognition. Two key components are fine-tuning the DNNs fully convolutionally with an objective function based on the saliency evaluation metrics, and integrating information at different image scales. We compare our method with 14 saliency models on 6 public eye tracking benchmark datasets. Results demonstrate that our DNNs can automatically learn features particularly for saliency prediction that surpass by a big margin the state-of-the-art. In addition, our model ranks top to date under all seven metrics on the MIT300 challenge set.

A significant body of literature on saliency modeling predicts where humans look in a single image or video. Besides the scientific goal of understanding how information is fused from multiple visual sources to identify regions of interest in a holistic manner, there are tremendous engineering applications of multi-camera saliency due to the widespread of cameras. This paper proposes a principled framework to smoothly integrate visual information from multiple views to a global scene map, and to employ a saliency algorithm incorporating high-level features to identify the most important regions by fusing visual information. The proposed method has the following key distinguishing features compared with its counterparts: (1) the proposed saliency detection is global (salient regions from one local view may not be important in a global context), (2) it does not require special ways for camera deployment or overlapping field of view, and (3) the key saliency algorithm is effective in highlighting interesting object regions though not a single detector is used. Experiments on several data sets confirm the effectiveness of the proposed principled framework.

This paper proposes to learn a discriminative dictionary for saliency detection. In addition to the conventional sparse coding mechanism that learns a representational dictionary of natural images for saliency prediction, this work uses supervised information from eye tracking experiments in training to enhance the discriminative power of the learned dictionary. Furthermore, we explicitly model saliency at multi-scale by formulating it as a multi-class problem, and a label consistency term is incorporated into the framework to encourage class (salient vs. non-salient) and scale consistency in the learned sparse codes. K-SVD is employed as the central computational module to efficiently obtain the optimal solution. Experiments demonstrate the superior performance of the proposed algorithm compared with the state-of-the-art in saliency prediction.

Studying signal and noise properties of recorded neural data is critical in developing more efficient algorithms to recover the encoded information. Important issues exist in this research including the variant spectrum spans of neural spikes that make it difficult to choose a global optimal bandpass filter. Also, multiple sources produce aggregated noise that deviates from the conventional white Gaussian noise. In this work, the spectrum variability of spikes is addressed, based on which the concept of adaptive bandpass filter that fits the spectrum of individual spikes is proposed. Multiple noise sources have been studied through analytical models as well as empirical measurements. The dominant noise source is identified as neuron noise followed by interface noise of the electrode. This suggests that major efforts to reduce noise from electronics are not well spent. The measured noise from in vivo experiments shows a family of 1/f^{x} (x=1.5\pm 0.5) spectrum that can be reduced using noise shaping techniques. In summary, the methods of adaptive bandpass filtering and noise shaping together result in several dB signal-to-noise ratio (SNR) enhancement.