This paper presents an end-to-end differentiable algorithm for robust and detail-preserving surface normal estimation on unstructured point-clouds. We utilize graph neural networks to iteratively parameterize an adaptive anisotropic kernel that produces point weights for weighted least-squares plane fitting in local neighborhoods. The approach retains the interpretability and efficiency of traditional sequential plane fitting while benefiting from adaptation to data set statistics through deep learning. This results in a state-of-the-art surface normal estimator that is robust to noise, outliers and point density variation, preserves sharp features through anisotropic kernels and equivariance through a local quaternion-based spatial transformer. Contrary to previous deep learning methods, the proposed approach does not require any hand-crafted features or preprocessing. It improves on the state-of-the-art results while being more than two orders of magnitude faster and more parameter efficient.

This paper introduces a neural style transfer model to generate a stylized image conditioning on a set of examples describing the desired style. The proposed solution produces high-quality images even in the zero-shot setting and allows for more freedom in changes to the content geometry. This is made possible by introducing a novel Two-Stage Peer-Regularization Layer that recombines style and content in latent space by means of a custom graph convolutional layer. Contrary to the vast majority of existing solutions, our model does not depend on any pre-trained networks for computing perceptual losses and can be trained fully end-to-end thanks to a new set of cyclic losses that operate directly in latent space and not on the RGB images. An extensive ablation study confirms the usefulness of the proposed losses and of the Two-Stage Peer-Regularization Layer, with qualitative results that are competitive with respect to the current state of the art using a single model for all presented styles. This opens the door to more abstract and artistic neural image generation scenarios, along with simpler deployment of the model.

The research in biometric recognition using hand shape has been somewhat stagnating in the last decade. Meanwhile, computer vision and machine learning have experienced a paradigm shift with the renaissance of deep learning, which has set the new state-of-the-art in many related fields. Inspired by successful applications of deep learning for other biometric modalities, we propose a novel approach to 3D hand shape recognition from RGB-D data based on geometric deep learning techniques. We show how to train our model on synthetic data and retain the performance on real samples during test time. To evaluate our method, we provide a new dataset NNHand RGB- D of short video sequences and show encouraging performance compared to diverse baselines on the new data, as well as current benchmark dataset HKPolyU. Moreover, the new dataset opens door to many new research directions in hand shape recognition.

Event cameras encode visual information with high temporal precision, low data-rate, and high-dynamic range. Thanks to these characteristics, event cameras are particularly suited for scenarios with high motion, challenging lighting conditions and requiring low latency. However, due to the novelty of the field, the performance of event-based systems on many vision tasks is still lower compared to conventional frame-based solutions. The main reasons for this performance gap are: the lower spatial resolution of event sensors, compared to frame cameras; the lack of large-scale training datasets; the absence of well established deep learning architectures for event-based processing. In this paper, we address all these problems in the context of an event-based object detection task. First, we publicly release the first high-resolution large-scale dataset for object detection. The dataset contains more than 14 hours recordings of a 1 megapixel event camera, in automotive scenarios, together with 25M bounding boxes of cars, pedestrians, and two-wheelers, labeled at high frequency. Second, we introduce a novel recurrent architecture for event-based detection and a temporal consistency loss for better-behaved training. The ability to compactly represent the sequence of events into the internal memory of the model is essential to achieve high accuracy. Our model outperforms by a large margin feed-forward event-based architectures. Moreover, our method does not require any reconstruction of intensity images from events, showing that training directly from raw events is possible, more efficient, and more accurate than passing through an intermediate intensity image. Experiments on the dataset introduced in this work, for which events and gray level images are available, show performance on par with that of highly tuned and studied frame-based detectors.

This paper introduces Non-Autonomous Input-Output Stable Network (NAIS-Net), a very deep architecture where each stacked processing block is derived from a time-invariant non-autonomous dynamical system. Non-autonomy is implemented by skip connections from the block input to each of the unrolled processing stages and allows stability to be enforced so that blocks can be unrolled adaptively to a pattern-dependent processing depth. NAIS-Net induces non-trivial, Lipschitz input-output maps, even for an infinite unroll length. We prove that the network is globally asymptotically stable so that for every initial condition there is exactly one input-dependent equilibrium assuming tanh units, and multiple stable equilibria for ReL units. An efficient implementation that enforces the stability under derived conditions for both fully-connected and convolutional layers is also presented. Experimental results show how NAIS-Net exhibits stability in practice, yielding a significant reduction in generalization gap compared to ResNets.

Deep learning has achieved a remarkable performance breakthrough in several fields, most notably in speech recognition, natural language processing, and computer vision. In particular, convolutional neural network (CNN) architectures currently produce state-of-the-art performance on a variety of image analysis tasks such as object detection and recognition. Most of deep learning research has so far focused on dealing with 1D, 2D, or 3D Euclidean-structured data such as acoustic signals, images, or videos. Recently, there has been an increasing interest in geometric deep learning, attempting to generalize deep learning methods to non-Euclidean structured data such as graphs and manifolds, with a variety of applications from the domains of network analysis, computational social science, or computer graphics. In this paper, we propose a unified framework allowing to generalize CNN architectures to non-Euclidean domains (graphs and manifolds) and learn local, stationary, and compositional task-specific features. We show that various non-Euclidean CNN methods previously proposed in the literature can be considered as particular instances of our framework. We test the proposed method on standard tasks from the realms of image-, graph-and 3D shape analysis and show that it consistently outperforms previous approaches.

We consider the problem of deformable object detection and dense correspondence in cluttered 3D scenes. Key ingredient to our method is the choice of representation: we formulate the problem in the spectral domain using the functional maps framework, where we seek for the most regular nearly-isometric parts in the model and the scene that minimize correspondence error. The problem is initialized by solving a sparse relaxation of a quadratic assignment problem on features obtained via data-driven metric learning. The resulting matching pipeline is solved efficiently, and yields accurate results in challenging settings that were previously left unexplored in the literature.

In the past years, deep convolutional neural networks (CNNs) have become extremely popular in the computer vision and pattern recognition community. The computational power of modern processors, efficient stochastic optimization algorithms, and large amounts of training data allowed training complex tasks-specific features directly from the data in an end-to-end fashion, as opposed to the traditional way of using hand-crafted feature descriptors. CNNs are currently state-of-the-art methods in many computer vision problems, and have been successfully used in biometric applications such as face, fingerpring, and voice recognition. In palmprint recognition applications, CNNs have not yet been explored, and the majority of methods still rely on hand-crafted representations which do not scale well to large datasets and that usually require a complex manual parameter tuning. In this work, we show that CNNs can be successfully used for palmprint recognition. The training of our network uses a novel loss function related to the d-prime index, which allows to achieve a better genuine/impostor score distribution separation than previous approaches with only little training data required. Our approach does not require cumbersome parameter tuning and achieves state-of-the-art results on the standard IIT Delhi and CASIA palmprint datasets.

Convolutional neural networks have achieved extraordinary results in many computer vision and pattern recognition applications; however, their adoption in the computer graphics and geometry processing communities is limited due to the non-Euclidean structure of their data. In this paper, we propose Anisotropic Convolutional Neural Network (ACNN), a generalization of classical CNNs to non-Euclidean domains, where classical convolutions are replaced by projections over a set of oriented anisotropic diffusion kernels. We use ACNNs to effectively learn intrinsic dense correspondences between deformable shapes, a fundamental problem in geometry processing, arising in a wide variety of applications. We tested ACNNs performance in very challenging settings, achieving state-of-the-art results on some of the most difficult recent correspondence benchmarks.

Abstract Spectral methods have recently gained popularity in many domains of computer graphics and geometry processing, especially shape processing, computation of shape descriptors, distances, and correspondence. Spectral geometric structures are intrinsic and thus invariant to isometric deformations, are efficiently computed, and can be constructed on shapes in different representations. A notable drawback of these constructions, however, is that they areisotropic, i.e., insensitive to direction. In this paper, we show how to construct direction‐sensitive spectral feature descriptors usinganisotropic diffusionon meshes and point clouds. The core of our construction are directed local kernels acting similarly to steerable filters, which are learned in a task‐specific manner. Remarkably, while being intrinsic, our descriptors allow to disambiguate reflection symmetries. We show the application of anisotropic descriptors for problems of shape correspondence on meshes and point clouds, achieving results significantly better than state‐of‐the‐art methods.

Abstract In this paper, we propose a generalization of convolutional neural networks (CNN) to non‐Euclidean domains for the analysis of deformable shapes. Our construction is based on localized frequency analysis (a generalization of the windowed Fourier transform to manifolds) that is used to extract the local behavior of some dense intrinsic descriptor, roughly acting as an analogy to patches in images. The resulting local frequency representations are then passed through a bank of filters whose coefficient are determined by a learning procedure minimizing a task‐specific cost. Our approach generalizes several previous methods such as HKS, WKS, spectral CNN, and GPS embeddings. Experimental results show that the proposed approach allows learning class‐specific shape descriptors significantly outperforming recent state‐of‐the‐art methods on standard benchmarks.

We introduce an efficient computational framework for hashing data belonging to multiple modalities into a single representation space where they become mutually comparable. The proposed approach is based on a novel coupled siamese neural network architecture and allows unified treatment of intra- and inter-modality similarity learning. Unlike existing cross-modality similarity learning approaches, our hashing functions are not limited to binarized linear projections and can assume arbitrarily complex forms. We show experimentally that our method significantly outperforms state-of-the-art hashing approaches on multimedia retrieval tasks.

Deep Neural Networks now excel at image classification, detection and segmentation. When used to scan images by means of a sliding window, however, their high computational complexity can bring even the most powerful hardware to its knees. We show how dynamic programming can speedup the process by orders of magnitude, even when max-pooling layers are present.

We present a practical, parameter-free, general computational-statistical technique for quantitative analysis of 2D images representing fiber-like structures (vessels, neurons, elongated objects, cell boundaries...), which is a common task in many experimental biomedicine scenarios. Our approach does not require segmentation or tracing of fibers; instead, it relies on a learned detector of intersections between fibers and arbitrary segments. The detector's probabilistic outputs are used to compute an estimate of the density of fibers and of its uncertainty; the latter accounts for several factors, including the intrinsic difficulty of the problem, i.e. the inaccuracy of the detector. After few minutes of training by the user, the procedure performs well in a variety of challenging scenarios, and compares favorably even with problem-specific algorithms.

We present a fast algorithm for training MaxPooling Convolutional Networks to segment images. This type of network yields record-breaking performance in a variety of tasks, but is normally trained on a computationally expensive patch-by-patch basis. Our new method processes each training image in a single pass, which is vastly more efficient. We validate the approach in different scenarios and report a 1500-fold speed-up. In an application to automated steel defect detection and segmentation, we obtain excellent performance with short training times.

We introduce a Multi-Scale Pyramidal Pooling Network tailored to generic steel defect classification, featuring a novel pyramidal pooling layer at multiple scales and a novel encoding layer. Thanks to the former, the network does not require all images of a given classification task to be of equal size. The latter narrows the gap to bag-of-features approaches. On various benchmark datasets, we evaluate and compare our system to convolutional neural networks and state-of-the-art computer vision methods. We also present results on a real industrial steel defect classification problem, where existing architectures are not applicable as they require equally sized input images. Our method substantially outperforms previous methods based on engineered features. It can be seen as a fully supervised hierarchical bag-of-features extension that is trained online and can be fine-tuned for any given task.

Local competition among neighboring neurons is common in biological neural networks (NNs). We apply the concept to gradient-based, backprop-trained artificial multilayer NNs. NNs with competing linear units tend to outperform those with non-competing nonlinear units, and avoid catastrophic forgetting when training sets change over time.

We present a Max-Pooling Convolutional Neural Network approach for supervised steel defect classification. On a classification task with 7 defects, collected from a real production line, an error rate of 7% is obtained. Compared to SVM classifiers trained on commonly used feature descriptors our best net performs at least two times better. Not only we do obtain much better results, but the proposed method also works directly on raw pixel intensities of detected and segmented steel defects, avoiding further time consuming and hard to optimize ad-hoc preprocessing.

We present a fast, fully parameterizable GPU implementation of Convolutional Neural Network variants. Our feature extractors are neither carefully designed nor pre-wired, but rather learned in a supervised way. Our deep hierarchical architectures achieve the best published results on benchmarks for object classification (NORB, CIFAR10) and handwritten digit recognition (MNIST), with error rates of 2.53%, 19.51%, 0.35%, respectively. Deep nets trained by simple back-propagation perform better than more shallow ones. Learning is surprisingly rapid. NORB is completely trained within five epochs. Test error rates on MNIST drop to 2.42%, 0.97 % and 0.48 % after 1, 3 and 17 epochs, respectively.

We describe the approach that won the preliminary phase of the German traffic sign recognition benchmark with a better-than-human recognition rate of 98.98%.We obtain an even better recognition rate of 99.15% by further training the nets. Our fast, fully parameterizable GPU implementation of a Convolutional Neural Network does not require careful design of pre-wired feature extractors, which are rather learned in a supervised way. A CNN/MLP committee further boosts recognition performance.