Sina Honari

Krzysztof Lis, Sina Honari, Pascal Fua et al.

While road obstacle detection techniques have become increasingly effective, they typically ignore the fact that, in practice, the apparent size of the obstacles decreases as their distance to the vehicle increases. In this paper, we account for this by computing a scale map encoding the apparent size of a hypothetical object at every image location. We then leverage this perspective map to (i) generate training data by injecting onto the road synthetic objects whose size corresponds to the perspective foreshortening; and (ii) incorporate perspective information in the decoding part of the detection network to guide the obstacle detector. Our results on standard benchmarks show that, together, these two strategies significantly boost the obstacle detection performance, allowing our approach to consistently outperform state-of-the-art methods in terms of instance-level obstacle detection.

Christopher Beckham, Martin Weiss, Florian Golemo et al.

Different types of mental rotation tests have been used extensively in psychology to understand human visual reasoning and perception. Understanding what an object or visual scene would look like from another viewpoint is a challenging problem that is made even harder if it must be performed from a single image. We explore a controlled setting whereby questions are posed about the properties of a scene if that scene was observed from another viewpoint. To do this we have created a new version of the CLEVR dataset that we call CLEVR Mental Rotation Tests (CLEVR-MRT). Using CLEVR-MRT we examine standard methods, show how they fall short, then explore novel neural architectures that involve inferring volumetric representations of a scene. These volumes can be manipulated via camera-conditioned transformations to answer the question. We examine the efficacy of different model variants through rigorous ablations and demonstrate the efficacy of volumetric representations.

Sina Honari, Chen Zhao, Mathieu Salzmann et al.

Analyzing and training 3D body posture models depend heavily on the availability of joint labels that are commonly acquired through laborious manual annotation of body joints or via marker-based joint localization using carefully curated markers and capturing systems. However, such annotations are not always available, especially for people performing unusual activities. In this paper, we propose an algorithm that learns to discover 3D keypoints on human bodies from multiple-view images without any supervision or labels other than the constraints multiple-view geometry provides. To ensure that the discovered 3D keypoints are meaningful, they are re-projected to each view to estimate the person's mask that the model itself has initially estimated without supervision. Our approach discovers more interpretable and accurate 3D keypoints compared to other state-of-the-art unsupervised approaches on Human3.6M and MPI-INF-3DHP benchmark datasets.

Amirhossein Kazerouni, Maitreya Suin, Tristan Aumentado-Armstrong et al.

Recent advances in image restoration have enabled high-fidelity recovery of faces from degraded inputs using reference-based face restoration models (Ref-FR). However, such methods focus solely on facial regions, neglecting degradation across the full scene, including body and background, which limits practical usability. Meanwhile, full-scene restorers often ignore degradation cues entirely, leading to underdetermined predictions and visual artifacts. In this work, we propose Face2Scene, a two-stage restoration framework that leverages the face as a perceptual oracle to estimate degradation and guide the restoration of the entire image. Given a degraded image and one or more identity references, we first apply a Ref-FR model to reconstruct high-quality facial details. From the restored-degraded face pair, we extract a face-derived degradation code that captures degradation attributes (e.g., noise, blur, compression), which is then transformed into multi-scale degradation-aware tokens. These tokens condition a diffusion model to restore the full scene in a single step, including the body and background. Extensive experiments demonstrate the superior effectiveness of the proposed method compared to state-of-the-art methods.

Krzysztof Lis, Matthias Rottmann, Annika Mütze et al.

In addition to impressive performance, vision transformers have demonstrated remarkable abilities to encode information they were not trained to extract. For example, this information can be used to perform segmentation or single-view depth estimation even though the networks were only trained for image recognition. We show that a similar phenomenon occurs when explicitly training transformers for semantic segmentation in a supervised manner for a set of categories: Once trained, they provide valuable information even about categories absent from the training set. This information can be used to segment objects from these never-seen-before classes in domains as varied as road obstacles, aircraft parked at a terminal, lunar rocks, and maritime hazards.

Soumava Kumar Roy, Ilia Badanin, Sina Honari et al.

Occlusions remain one of the key challenges in 3D body pose estimation from single-camera video sequences. Temporal consistency has been extensively used to mitigate their impact but the existing algorithms in the literature do not explicitly model them. Here, we apply this by representing the deforming body as a spatio-temporal graph. We then introduce a refinement network that performs graph convolutions over this graph to output 3D poses. To ensure robustness to occlusions, we train this network with a set of binary masks that we use to disable some of the edges as in drop-out techniques. In effect, we simulate the fact that some joints can be hidden for periods of time and train the network to be immune to that. We demonstrate the effectiveness of this approach compared to state-of-the-art techniques that infer poses from single-camera sequences.

Soumava Kumar Roy, Leonardo Citraro, Sina Honari et al.

Supervised approaches to 3D pose estimation from single images are remarkably effective when labeled data is abundant. However, as the acquisition of ground-truth 3D labels is labor intensive and time consuming, recent attention has shifted towards semi- and weakly-supervised learning. Generating an effective form of supervision with little annotations still poses major challenge in crowded scenes. In this paper we propose to impose multi-view geometrical constraints by means of a weighted differentiable triangulation and use it as a form of self-supervision when no labels are available. We therefore train a 2D pose estimator in such a way that its predictions correspond to the re-projection of the triangulated 3D pose and train an auxiliary network on them to produce the final 3D poses. We complement the triangulation with a weighting mechanism that alleviates the impact of noisy predictions caused by self-occlusion or occlusion from other subjects. We demonstrate the effectiveness of our semi-supervised approach on Human3.6M and MPI-INF-3DHP datasets, as well as on a new multi-view multi-person dataset that features occlusion.

Andrey Davydov, Anastasia Remizova, Victor Constantin et al.

The Skinned Multi-Person Linear (SMPL) model can represent a human body by mapping pose and shape parameters to body meshes. This has been shown to facilitate inferring 3D human pose and shape from images via different learning models. However, not all pose and shape parameter values yield physically-plausible or even realistic body meshes. In other words, SMPL is under-constrained and may thus lead to invalid results when used to reconstruct humans from images, either by directly optimizing its parameters, or by learning a mapping from the image to these parameters. In this paper, we therefore learn a prior that restricts the SMPL parameters to values that produce realistic poses via adversarial training. We show that our learned prior covers the diversity of the real-data distribution, facilitates optimization for 3D reconstruction from 2D keypoints, and yields better pose estimates when used for regression from images. We found that the prior based on spherical distribution gets the best results. Furthermore, in all these tasks, it outperforms the state-of-the-art VAE-based approach to constraining the SMPL parameters.

Semih Günel, Florian Aymanns, Sina Honari et al.

Relating animal behaviors to brain activity is a fundamental goal in neuroscience, with practical applications in building robust brain-machine interfaces. However, the domain gap between individuals is a major issue that prevents the training of general models that work on unlabeled subjects. Since 3D pose data can now be reliably extracted from multi-view video sequences without manual intervention, we propose to use it to guide the encoding of neural action representations together with a set of neural and behavioral augmentations exploiting the properties of microscopy imaging. To reduce the domain gap, during training, we swap neural and behavioral data across animals that seem to be performing similar actions. To demonstrate this, we test our methods on three very different multimodal datasets; one that features flies and their neural activity, one that contains human neural Electrocorticography (ECoG) data, and lastly the RGB video data of human activities from different viewpoints.

Semih Günel, Florian Aymanns, Sina Honari et al.

A fundamental goal in neuroscience is to understand the relationship between neural activity and behavior. For example, the ability to extract behavioral intentions from neural data, or neural decoding, is critical for developing effective brain machine interfaces. Although simple linear models have been applied to this challenge, they cannot identify important non-linear relationships. Thus, a self-supervised means of identifying non-linear relationships between neural dynamics and behavior, in order to compute neural representations, remains an important open problem. To address this challenge, we generated a new multimodal dataset consisting of the spontaneous behaviors generated by fruit flies, Drosophila melanogaster -- a popular model organism in neuroscience research. The dataset includes 3D markerless motion capture data from six camera views of the animal generating spontaneous actions, as well as synchronously acquired two-photon microscope images capturing the activity of descending neuron populations that are thought to drive actions. Standard contrastive learning and unsupervised domain adaptation techniques struggle to learn neural action representations (embeddings computed from the neural data describing action labels) due to large inter-animal differences in both neural and behavioral modalities. To overcome this deficiency, we developed simple yet effective augmentations that close the inter-animal domain gap, allowing us to extract behaviorally relevant, yet domain agnostic, information from neural data. This multimodal dataset and our new set of augmentations promise to accelerate the application of self-supervised learning methods in neuroscience.

Robin Chan, Krzysztof Lis, Svenja Uhlemeyer et al.

State-of-the-art semantic or instance segmentation deep neural networks (DNNs) are usually trained on a closed set of semantic classes. As such, they are ill-equipped to handle previously-unseen objects. However, detecting and localizing such objects is crucial for safety-critical applications such as perception for automated driving, especially if they appear on the road ahead. While some methods have tackled the tasks of anomalous or out-of-distribution object segmentation, progress remains slow, in large part due to the lack of solid benchmarks; existing datasets either consist of synthetic data, or suffer from label inconsistencies. In this paper, we bridge this gap by introducing the "SegmentMeIfYouCan" benchmark. Our benchmark addresses two tasks: Anomalous object segmentation, which considers any previously-unseen object category; and road obstacle segmentation, which focuses on any object on the road, may it be known or unknown. We provide two corresponding datasets together with a test suite performing an in-depth method analysis, considering both established pixel-wise performance metrics and recent component-wise ones, which are insensitive to object sizes. We empirically evaluate multiple state-of-the-art baseline methods, including several models specifically designed for anomaly / obstacle segmentation, on our datasets and on public ones, using our test suite. The anomaly and obstacle segmentation results show that our datasets contribute to the diversity and difficulty of both data landscapes.

Krzysztof Lis, Sina Honari, Pascal Fua et al.

Vehicles can encounter a myriad of obstacles on the road, and it is impossible to record them all beforehand to train a detector. Instead, we select image patches and inpaint them with the surrounding road texture, which tends to remove obstacles from those patches. We then use a network trained to recognize discrepancies between the original patch and the inpainted one, which signals an erased obstacle.

Sina Honari, Victor Constantin, Helge Rhodin et al.

In this paper we propose an unsupervised feature extraction method to capture temporal information on monocular videos, where we detect and encode subject of interest in each frame and leverage contrastive self-supervised (CSS) learning to extract rich latent vectors. Instead of simply treating the latent features of nearby frames as positive pairs and those of temporally-distant ones as negative pairs as in other CSS approaches, we explicitly disentangle each latent vector into a time-variant component and a time-invariant one. We then show that applying contrastive loss only to the time-variant features and encouraging a gradual transition on them between nearby and away frames while also reconstructing the input, extract rich temporal features, well-suited for human pose estimation. Our approach reduces error by about 50% compared to the standard CSS strategies, outperforms other unsupervised single-view methods and matches the performance of multi-view techniques. When 2D pose is available, our approach can extract even richer latent features and improve the 3D pose estimation accuracy, outperforming other state-of-the-art weakly supervised methods.

Edoardo Remelli, Shangchen Han, Sina Honari et al.

We present a lightweight solution to recover 3D pose from multi-view images captured with spatially calibrated cameras. Building upon recent advances in interpretable representation learning, we exploit 3D geometry to fuse input images into a unified latent representation of pose, which is disentangled from camera view-points. This allows us to reason effectively about 3D pose across different views without using compute-intensive volumetric grids. Our architecture then conditions the learned representation on camera projection operators to produce accurate per-view 2d detections, that can be simply lifted to 3D via a differentiable Direct Linear Transform (DLT) layer. In order to do it efficiently, we propose a novel implementation of DLT that is orders of magnitude faster on GPU architectures than standard SVD-based triangulation methods. We evaluate our approach on two large-scale human pose datasets (H36M and Total Capture): our method outperforms or performs comparably to the state-of-the-art volumetric methods, while, unlike them, yielding real-time performance.

MohammadHossein AskariHemmat, Sina Honari, Lucas Rouhier et al.

Model quantization is leveraged to reduce the memory consumption and the computation time of deep neural networks. This is achieved by representing weights and activations with a lower bit resolution when compared to their high precision floating point counterparts. The suitable level of quantization is directly related to the model performance. Lowering the quantization precision (e.g. 2 bits), reduces the amount of memory required to store model parameters and the amount of logic required to implement computational blocks, which contributes to reducing the power consumption of the entire system. These benefits typically come at the cost of reduced accuracy. The main challenge is to quantize a network as much as possible, while maintaining the performance accuracy. In this work, we present a quantization method for the U-Net architecture, a popular model in medical image segmentation. We then apply our quantization algorithm to three datasets: (1) the Spinal Cord Gray Matter Segmentation (GM), (2) the ISBI challenge for segmentation of neuronal structures in Electron Microscopic (EM), and (3) the public National Institute of Health (NIH) dataset for pancreas segmentation in abdominal CT scans. The reported results demonstrate that with only 4 bits for weights and 6 bits for activations, we obtain 8 fold reduction in memory requirements while loosing only 2.21%, 0.57% and 2.09% dice overlap score for EM, GM and NIH datasets respectively. Our fixed point quantization provides a flexible trade off between accuracy and memory requirement which is not provided by previous quantization methods for U-Net such as TernaryNet.

Christopher Beckham, Sina Honari, Vikas Verma et al.

In this paper, we explore new approaches to combining information encoded within the learned representations of auto-encoders. We explore models that are capable of combining the attributes of multiple inputs such that a resynthesised output is trained to fool an adversarial discriminator for real versus synthesised data. Furthermore, we explore the use of such an architecture in the context of semi-supervised learning, where we learn a mixing function whose objective is to produce interpolations of hidden states, or masked combinations of latent representations that are consistent with a conditioned class label. We show quantitative and qualitative evidence that such a formulation is an interesting avenue of research.

Joseph Paul Cohen, Margaux Luck, Sina Honari

This paper discusses how distribution matching losses, such as those used in CycleGAN, when used to synthesize medical images can lead to mis-diagnosis of medical conditions. It seems appealing to use these new image synthesis methods for translating images from a source to a target domain because they can produce high quality images and some even do not require paired data. However, the basis of how these image translation models work is through matching the translation output to the distribution of the target domain. This can cause an issue when the data provided in the target domain has an over or under representation of some classes (e.g. healthy or sick). When the output of an algorithm is a transformed image there are uncertainties whether all known and unknown class labels have been preserved or changed. Therefore, we recommend that these translated images should not be used for direct interpretation (e.g. by doctors) because they may lead to misdiagnosis of patients based on hallucinated image features by an algorithm that matches a distribution. However there are many recent papers that seem as though this is the goal.

Joel Ruben Antony Moniz, Christopher Beckham, Simon Rajotte et al.

We present an unsupervised approach for learning to estimate three dimensional (3D) facial structure from a single image while also predicting 3D viewpoint transformations that match a desired pose and facial geometry. We achieve this by inferring the depth of facial keypoints of an input image in an unsupervised manner, without using any form of ground-truth depth information. We show how it is possible to use these depths as intermediate computations within a new backpropable loss to predict the parameters of a 3D affine transformation matrix that maps inferred 3D keypoints of an input face to the corresponding 2D keypoints on a desired target facial geometry or pose. Our resulting approach, called DepthNets, can therefore be used to infer plausible 3D transformations from one face pose to another, allowing faces to be frontalized, transformed into 3D models or even warped to another pose and facial geometry. Lastly, we identify certain shortcomings with our formulation, and explore adversarial image translation techniques as a post-processing step to re-synthesize complete head shots for faces re-targeted to different poses or identities.

Shanxin Yuan, Guillermo Garcia-Hernando, Bjorn Stenger et al.

In this paper, we strive to answer two questions: What is the current state of 3D hand pose estimation from depth images? And, what are the next challenges that need to be tackled? Following the successful Hands In the Million Challenge (HIM2017), we investigate the top 10 state-of-the-art methods on three tasks: single frame 3D pose estimation, 3D hand tracking, and hand pose estimation during object interaction. We analyze the performance of different CNN structures with regard to hand shape, joint visibility, view point and articulation distributions. Our findings include: (1) isolated 3D hand pose estimation achieves low mean errors (10 mm) in the view point range of [70, 120] degrees, but it is far from being solved for extreme view points; (2) 3D volumetric representations outperform 2D CNNs, better capturing the spatial structure of the depth data; (3) Discriminative methods still generalize poorly to unseen hand shapes; (4) While joint occlusions pose a challenge for most methods, explicit modeling of structure constraints can significantly narrow the gap between errors on visible and occluded joints.

Sina Honari, Pavlo Molchanov, Stephen Tyree et al.

We present two techniques to improve landmark localization in images from partially annotated datasets. Our primary goal is to leverage the common situation where precise landmark locations are only provided for a small data subset, but where class labels for classification or regression tasks related to the landmarks are more abundantly available. First, we propose the framework of sequential multitasking and explore it here through an architecture for landmark localization where training with class labels acts as an auxiliary signal to guide the landmark localization on unlabeled data. A key aspect of our approach is that errors can be backpropagated through a complete landmark localization model. Second, we propose and explore an unsupervised learning technique for landmark localization based on having a model predict equivariant landmarks with respect to transformations applied to the image. We show that these techniques, improve landmark prediction considerably and can learn effective detectors even when only a small fraction of the dataset has landmark labels. We present results on two toy datasets and four real datasets, with hands and faces, and report new state-of-the-art on two datasets in the wild, e.g. with only 5\% of labeled images we outperform previous state-of-the-art trained on the AFLW dataset.

Florian Bordes, Sina Honari, Pascal Vincent

In this work, we investigate a novel training procedure to learn a generative model as the transition operator of a Markov chain, such that, when applied repeatedly on an unstructured random noise sample, it will denoise it into a sample that matches the target distribution from the training set. The novel training procedure to learn this progressive denoising operation involves sampling from a slightly different chain than the model chain used for generation in the absence of a denoising target. In the training chain we infuse information from the training target example that we would like the chains to reach with a high probability. The thus learned transition operator is able to produce quality and varied samples in a small number of steps. Experiments show competitive results compared to the samples generated with a basic Generative Adversarial Net

The Theano Development Team, Rami Al-Rfou, Guillaume Alain et al.

Theano is a Python library that allows to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. Since its introduction, it has been one of the most used CPU and GPU mathematical compilers - especially in the machine learning community - and has shown steady performance improvements. Theano is being actively and continuously developed since 2008, multiple frameworks have been built on top of it and it has been used to produce many state-of-the-art machine learning models. The present article is structured as follows. Section I provides an overview of the Theano software and its community. Section II presents the principal features of Theano and how to use them, and compares them with other similar projects. Section III focuses on recently-introduced functionalities and improvements. Section IV compares the performance of Theano against Torch7 and TensorFlow on several machine learning models. Section V discusses current limitations of Theano and potential ways of improving it.

David Rim, Sina Honari, Md Kamrul Hasan et al.

We present techniques for improving performance driven facial animation, emotion recognition, and facial key-point or landmark prediction using learned identity invariant representations. Established approaches to these problems can work well if sufficient examples and labels for a particular identity are available and factors of variation are highly controlled. However, labeled examples of facial expressions, emotions and key-points for new individuals are difficult and costly to obtain. In this paper we improve the ability of techniques to generalize to new and unseen individuals by explicitly modeling previously seen variations related to identity and expression. We use a weakly-supervised approach in which identity labels are used to learn the different factors of variation linked to identity separately from factors related to expression. We show how probabilistic modeling of these sources of variation allows one to learn identity-invariant representations for expressions which can then be used to identity-normalize various procedures for facial expression analysis and animation control. We also show how to extend the widely used techniques of active appearance models and constrained local models through replacing the underlying point distribution models which are typically constructed using principal component analysis with identity-expression factorized representations. We present a wide variety of experiments in which we consistently improve performance on emotion recognition, markerless performance-driven facial animation and facial key-point tracking.

Sina Honari, Jason Yosinski, Pascal Vincent et al.

Deep neural networks with alternating convolutional, max-pooling and decimation layers are widely used in state of the art architectures for computer vision. Max-pooling purposefully discards precise spatial information in order to create features that are more robust, and typically organized as lower resolution spatial feature maps. On some tasks, such as whole-image classification, max-pooling derived features are well suited; however, for tasks requiring precise localization, such as pixel level prediction and segmentation, max-pooling destroys exactly the information required to perform well. Precise localization may be preserved by shallow convnets without pooling but at the expense of robustness. Can we have our max-pooled multi-layered cake and eat it too? Several papers have proposed summation and concatenation based methods for combining upsampled coarse, abstract features with finer features to produce robust pixel level predictions. Here we introduce another model --- dubbed Recombinator Networks --- where coarse features inform finer features early in their formation such that finer features can make use of several layers of computation in deciding how to use coarse features. The model is trained once, end-to-end and performs better than summation-based architectures, reducing the error from the previous state of the art on two facial keypoint datasets, AFW and AFLW, by 30\% and beating the current state-of-the-art on 300W without using extra data. We improve performance even further by adding a denoising prediction model based on a novel convnet formulation.