CVPR 2019 Papers

Estimating hand-object manipulations is essential for in- terpreting and imitating human actions. Previous work has made significant progress towards reconstruction of hand poses and object shapes in isolation. Yet, reconstructing hands and objects during manipulation is a more challeng- ing task due to significant occlusions of both the hand and object. While presenting challenges, manipulations may also simplify the problem since the physics of contact re- stricts the space of valid hand-object configurations. For example, during manipulation, the hand and object should be in contact but not interpenetrate. In this work, we regu- larize the joint reconstruction of hands and objects with ma- nipulation constraints. We present an end-to-end learnable model that exploits a novel contact loss that favors phys- ically plausible hand-object constellations. Our approach improves grasp quality metrics over baselines, using RGB images as input. To train and evaluate the model, we also propose a new large-scale synthetic dataset, ObMan, with hand-object manipulations. We demonstrate the transfer- ability of ObMan-trained models to real data.

Gait recognition is an important biometric method popularly used in video surveillance, where the task is to identify people at a distance by their walking patterns from video sequences. Most of the current successful approaches for gait recognition either use a pair of gait images to form a cross-gait representation or rely on a single gait image for unique-gait representation. These two types of representations emperically complement one another. In this paper, we propose a new Joint Unique-gait and Cross-gait Network (JUCNet), to combine the advantages of unique-gait representation with that of cross-gait representation, leading to an significantly improved performance. Another key contribution of this paper is a novel quintuplet loss function, which simultaneously increases the inter-class differences by pushing representations extracted from different subjects apart and decreases the intra-class variations by pulling representations extracted from the same subject together. Experiments show that our method achieves the state-of-the-art performance tested on standard benchmark datasets, demonstrating its superiority over existing methods.

Discriminative region responses residing inside an object instance can be extracted from networks trained with image-level label supervision. However, learning the full extent of pixel-level instance response in a weakly supervised manner remains unexplored. In this work, we tackle this challenging problem by using a novel instance extent filling approach. We first design a process to selectively collect pseudo supervision from noisy segment proposals obtained with previously published techniques. The pseudo supervision is used to learn a differentiable filling module that predicts a class-agnostic activation map for each instance given the image and an incomplete region response. We refer to the above maps as Instance Activation Maps (IAMs), which provide a fine-grained instance-level representation and allow instance masks to be extracted by lightweight CRF. Extensive experiments on the PASCAL VOC12 dataset show that our approach beats the state-of-the-art weakly supervised instance segmentation methods by a significant margin and increases the inference speed by an order of magnitude. Our method also generalizes well across domains and to unseen object categories. Without fine-tuning for the specific tasks, our model trained on VOC12 dataset (20 classes) obtains top performance for weakly supervised object localization on the CUB dataset (200 classes) and achieves competitive results on three widely used salient object detection benchmarks.

Human speech is often accompanied by hand and arm gestures. We present a method for cross-modal translation from "in-the-wild" monologue speech of a single speaker to their conversational gesture motion. We train on unlabeled videos for which we only have noisy pseudo ground truth from an automatic pose detection system. Our proposed model significantly outperforms baseline methods in a quantitative comparison. To support research toward obtaining a computational understanding of the relationship between gesture and speech, we release a large video dataset of person-specific gestures.

We present a system for learning motion maps of independently moving objects from stereo videos. The only annotations used in our system are 2D object bounding boxes which introduce the notion of objects in our system. Unlike prior learning based approaches which have focused on predicting dense optical flow fields and/or depth maps for images, we propose to predict instance specific 3D scene flow maps and instance masks from which we derive a factored 3D motion map for each object instance. Our network takes the 3D geometry of the problem into account which allows it to correlate the input images and distinguish moving objects from static ones. We present experiments evaluating the accuracy of our 3D flow vectors, as well as depth maps and projected 2D optical flow where our jointly learned system outperforms earlier approaches trained for each task independently.

We advocate the use of implicit fields for learning generative models of shapes and introduce an implicit field decoder, called IM-NET, for shape generation, aimed at improving the visual quality of the generated shapes. An implicit field assigns a value to each point in 3D space, so that a shape can be extracted as an iso-surface. IM-NET is trained to perform this assignment by means of a binary classifier. Specifically, it takes a point coordinate, along with a feature vector encoding a shape, and outputs a value which indicates whether the point is outside the shape or not. By replacing conventional decoders by our implicit decoder for representation learning (via IM-AE) and shape generation (via IM-GAN), we demonstrate superior results for tasks such as generative shape modeling, interpolation, and single-view 3D reconstruction, particularly in terms of visual quality. Code and supplementary material are available at https://github.com/czq142857/implicit-decoder.

Compression has been an important research topic for many decades, to produce a significant impact on data transmission and storage. Recent advances have shown a great potential of learning based image and video compression. Inspired from related works, in this paper, we present an image compression architecture using a convolutional autoencoder, and then generalize image compression to video compression, by adding an interpolation loop into both encoder and decoder sides. Our basic idea is to realize spatial-temporal energy compaction in learning image and video compression. Thereby, we propose to add a spatial energy compaction-based penalty into loss function, to achieve higher image compression performance. Furthermore, based on temporal energy distribution, we propose to select the number of frames in one interpolation loop, adapting to the motion characteristics of video contents. Experimental results demonstrate that our proposed image compression outperforms the latest image compression standard with MS-SSIM quality metric, and provides higher performance compared with state-of-the-art learning compression methods at high bit rates, which benefits from our spatial energy compaction approach. Meanwhile, our proposed video compression approach with temporal energy compaction can significantly outperform MPEG-4, and is competitive with commonly used H.264. Both our image and video compression can produce more visually pleasant results than traditional standards.

Recently, counting the number of people for crowd scenes is a hot topic because of its widespread applications (e.g. video surveillance, public security). It is a difficult task in the wild: changeable environment, large-range number of people cause the current methods can not work well. In addition, due to the scarce data, many methods suffer from over-fitting to a different extent. To remedy the above two problems, firstly, we develop a data collector and labeler, which can generate the synthetic crowd scenes and simultaneously annotate them without any manpower. Based on it, we build a large-scale, diverse synthetic dataset. Secondly, we propose two schemes that exploit the synthetic data to boost the performance of crowd counting in the wild: 1) pretrain a crowd counter on the synthetic data, then finetune it using the real data, which significantly prompts the model's performance on real data; 2) propose a crowd counting method via domain adaptation, which can free humans from heavy data annotations. Extensive experiments show that the first method achieves the state-of-the-art performance on four real datasets, and the second outperforms our baselines. The dataset and source code are available at https://gjy3035.github.io/GCC-CL/.

The predictive performance of supervised learning algorithms depends on the quality of labels. In a typical label collection process, multiple annotators provide subjective noisy estimates of the "truth" under the influence of their varying skill-levels and biases. Blindly treating these noisy labels as the ground truth limits the accuracy of learning algorithms in the presence of strong disagreement. This problem is critical for applications in domains such as medical imaging where both the annotation cost and inter-observer variability are high. In this work, we present a method for simultaneously learning the individual annotator model and the underlying true label distribution, using only noisy observations. Each annotator is modeled by a confusion matrix that is jointly estimated along with the classifier predictions. We propose to add a regularization term to the loss function that encourages convergence to the true annotator confusion matrix. We provide a theoretical argument as to how the regularization is essential to our approach both for the case of single annotator and multiple annotators. Despite the simplicity of the idea, experiments on image classification tasks with both simulated and real labels show that our method either outperforms or performs on par with the state-of-the-art methods and is capable of estimating the skills of annotators even with a single label available per image.

Food computing is playing an increasingly important role in human daily life, and has found tremendous applications in guiding human behavior towards smart food consumption and healthy lifestyle. An important task under the food-computing umbrella is retrieval, which is particularly helpful for health related applications, where we are interested in retrieving important information about food (e.g., ingredients, nutrition, etc.). In this paper, we investigate an open research task of cross-modal retrieval between cooking recipes and food images, and propose a novel framework Adversarial Cross-Modal Embedding (ACME) to resolve the cross-modal retrieval task in food domains. Specifically, the goal is to learn a common embedding feature space between the two modalities, in which our approach consists of several novel ideas: (i) learning by using a new triplet loss scheme together with an effective sampling strategy, (ii) imposing modality alignment using an adversarial learning strategy, and (iii) imposing cross-modal translation consistency such that the embedding of one modality is able to recover some important information of corresponding instances in the other modality. ACME achieves the state-of-the-art performance on the benchmark Recipe1M dataset, validating the efficacy of the proposed technique.

We introduce a self-supervised method for learning visual correspondence from unlabeled video. The main idea is to use cycle-consistency in time as free supervisory signal for learning visual representations from scratch. At training time, our model learns a feature map representation to be useful for performing cycle-consistent tracking. At test time, we use the acquired representation to find nearest neighbors across space and time. We demonstrate the generalizability of the representation -- without finetuning -- across a range of visual correspondence tasks, including video object segmentation, keypoint tracking, and optical flow. Our approach outperforms previous self-supervised methods and performs competitively with strongly supervised methods.

Person re-identification has achieved great progress with deep convolutional neural networks. However, most previous methods focus on learning individual appearance feature embedding, and it is hard for the models to handle difficult situations with different illumination, large pose variance and occlusion. In this work, we take a step further and consider employing context information for person search. For a probe-gallery pair, we first propose a contextual instance expansion module, which employs a relative attention module to search and filter useful context information in the scene. We also build a graph learning framework to effectively employ context pairs to update target similarity. These two modules are built on top of a joint detection and instance feature learning framework, which improves the discriminativeness of the learned features. The proposed framework achieves state-of-the-art performance on two widely used person search datasets.

In this paper, we propose a channel-wise interaction based binary convolutional neural network learning method (CI-BCNN) for efficient inference. Conventional methods apply xnor and bitcount operations in binary convolution with notable quantization error, which usually obtains inconsistent signs in binary feature maps compared with their full-precision counterpart and leads to significant information loss. In contrast, our CI-BCNN mines the channel-wise interactions, through which prior knowledge is provided to alleviate inconsistency of signs in binary feature maps and preserves the information of input samples during inference. Specifically, we mine the channel-wise interactions by a reinforcement learning model, and impose channel-wise priors on the intermediate feature maps through the interacted bitcount function. Extensive experiments on the CIFAR-10 and ImageNet datasets show that our method outperforms the state-of-the-art binary convolutional neural networks with less computational and storage cost.

With the rapid growth of fashion-focused social networks and online shopping, intelligent fashion recommendation is now in great needs. Recommending fashion outfits, each of which is composed of multiple interacted clothing and accessories, is relatively new to the field. The problem becomes even more interesting and challenging when considering users' personalized fashion style. Another challenge in a large-scale fashion outfit recommendation system is the efficiency issue of item/outfit search and storage. In this paper, we propose to learn binary code for efficient personalized fashion outfits recommendation. Our system consists of three components, a feature network for content extraction, a set of type-dependent hashing modules to learn binary codes, and a matching block that conducts pairwise matching. The whole framework is trained in an end-to-end manner. We collect outfit data together with user label information from a fashion-focused social website for the personalized recommendation task. Extensive experiments on our datasets show that the proposed framework outperforms the state-of-the-art methods significantly even with a simple backbone.

This paper presents a region-partition based attraction field dual representation for line segment maps, and thus poses the problem of line segment detection (LSD) as the region coloring problem. The latter is then addressed by learning deep convolutional neural networks (ConvNets) for accuracy, robustness and efficiency. For a 2D line segment map, our dual representation consists of three components: (i) A region-partition map in which every pixel is assigned to one and only one line segment; (ii) An attraction field map in which every pixel in a partition region is encoded by its 2D projection vector w.r.t. the associated line segment; and (iii) A squeeze module which squashes the attraction field to a line segment map that almost perfectly recovers the input one. By leveraging the duality, we learn ConvNets to compute the attraction field maps for raw in-put images, followed by the squeeze module for LSD, in an end-to-end manner. Our method rigorously addresses several challenges in LSD such as local ambiguity and class imbalance. Our method also harnesses the best practices developed in ConvNets based semantic segmentation methods such as the encoder-decoder architecture and the a-trous convolution. In experiments, our method is tested on the WireFrame dataset and the YorkUrban dataset with state-of-the-art performance obtained. Especially, we advance the performance by 4.5 percents on the WireFramedataset. Our method is also fast with 6.6 10.4 FPS, outperforming most of existing line segment detectors.

Modeling relation between actors is important for recognizing group activity in a multi-person scene. This paper aims at learning discriminative relation between actors efficiently using deep models. To this end, we propose to build a flexible and efficient \rm Actor Relation Graph (ARG) to simultaneously capture the appearance and position relation between actors. Thanks to the Graph Convolutional Network, the connections in ARG could be automatically learned from group activity videos in an end-to-end manner, and the inference on ARG could be efficiently performed with standard matrix operations. Furthermore, in practice, we come up with two variants to sparsify ARG for more effective modeling in videos: spatially localized ARG and temporal randomized ARG. We perform extensive experiments on two standard group activity recognition datasets: the Volleyball dataset and the Collective Activity dataset, where state-of-the-art performance is achieved on both datasets. We also visualize the learned actor graphs and relation features, which demonstrate that the proposed ARG is able to capture the discriminative relation information for group activity recognition.

Image segmentation is an important step in medical image processing and has been widely studied and developed for refinement of clinical analysis and applications. New models based on deep learning have improved results but are restricted to pixel-wise fitting of the segmentation map. Our aim was to tackle this limitation by developing a new model based on deep learning which takes into account the area inside as well as outside the region of interest as well as the size of boundaries during learning. Specifically, we propose a new loss function which incorporates area and size information and integrates this into a dense deep learning model. We evaluated our approach on a dataset of more than 2,000 cardiac MRI scans. Our results show that the proposed loss function outperforms other mainstream loss function Cross-entropy on two common segmentation networks. Our loss function is robust while using different hyperparameter lambda.

From an image of a person in action, we can easily guess the 3D motion of the person in the immediate past and future. This is because we have a mental model of 3D human dynamics that we have acquired from observing visual sequences of humans in motion. We present a framework that can similarly learn a representation of 3D dynamics of humans from video via a simple but effective temporal encoding of image features. At test time, from video, the learned temporal representation give rise to smooth 3D mesh predictions. From a single image, our model can recover the current 3D mesh as well as its 3D past and future motion. Our approach is designed so it can learn from videos with 2D pose annotations in a semi-supervised manner. Though annotated data is always limited, there are millions of videos uploaded daily on the Internet. In this work, we harvest this Internet-scale source of unlabeled data by training our model on unlabeled video with pseudo-ground truth 2D pose obtained from an off-the-shelf 2D pose detector. Our experiments show that adding more videos with pseudo-ground truth 2D pose monotonically improves 3D prediction performance. We evaluate our model on the recent challenging dataset of 3D Poses in the Wild and obtain state-of-the-art performance on the 3D prediction task without any fine-tuning. The project website with video can be found at https://akanazawa.github.io/human_dynamics/.

Detecting dense landmarks for diverse clothes, as a fundamental technique for clothes analysis, has attracted increasing research attention due to its huge application potential. However, due to the lack of modeling underlying semantic layout constraints among landmarks, prior works often detect ambiguous and structure-inconsistent landmarks of multiple overlapped clothes in one person. In this paper, we propose to seamlessly enforce structural layout relationships among landmarks on the intermediate representations via multiple stacked layout-graph reasoning layers. We define the layout-graph as a hierarchical structure including a root node, body-part nodes (e.g. upper body, lower body), coarse clothes-part nodes (e.g. collar, sleeve) and leaf landmark nodes (e.g. left-collar, right-collar). Each Layout-Graph Reasoning(LGR) layer aims to map feature representations into structural graph nodes via a Map-to-Node module, performs reasoning over structural graph nodes to achieve global layout coherency via a layout-graph reasoning module, and then maps graph nodes back to enhance feature representations via a Node-to-Map module. The layout-graph reasoning module integrates a graph clustering operation to generate representations of intermediate nodes (bottom-up inference) and then a graph deconvolution operation (top-down inference) over the whole graph. Extensive experiments on two public fashion landmark datasets demonstrate the superiority of our model. Furthermore, to advance the fine-grained fashion landmark research for supporting more comprehensive clothes generation and attribute recognition, we contribute the first Fine-grained Fashion Landmark Dataset (FFLD) containing 200k images annotated with at most 32 key-points for 13 clothes types.