ICCV 2017 Papers

Video deblurring is a challenging problem as the blur is complex and usually caused by the combination of camera shakes, object motions, and depth variations. Optical flow can be used for kernel estimation since it predicts motion trajectories. However, the estimates are often inaccurate in complex scenes at object boundaries, which are crucial in kernel estimation. In this paper, we exploit semantic segmentation in each blurry frame to understand the scene contents and use different motion models for image regions to guide optical flow estimation. While existing pixel-wise blur models assume that the blur kernel is the same as optical flow during the exposure time, this assumption does not hold when the motion blur trajectory at a pixel is different from the estimated linear optical flow. We analyze the relationship between motion blur trajectory and optical flow, and present a novel pixel-wise non-linear kernel model to account for motion blur. The proposed blur model is based on the non-linear optical flow, which describes complex motion blur more effectively. Extensive experiments on challenging blurry videos demonstrate the proposed algorithm performs favorably against the state-of-the-art methods.

VegFru: A Domain-Specific Dataset for Fine-grained Visual Categorization In this paper, we propose a novel domain-specific dataset named VegFru for fine-grained visual categorization (FGVC). While the existing datasets for FGVC are mainly focused on animal breeds or man-made objects with limited labelled data, VegFru is a larger dataset consisting of vegetables and fruits which are closely associated with the daily life of everyone. Aiming at domestic cooking and food management, VegFru categorizes vegetables and fruits according to their eating characteristics, and each image contains at least one edible part of vegetables or fruits with the same cooking usage. Particularly, all the images are labelled hierarchically. The current version covers vegetables and fruits of 25 upper-level categories and 292 subordinate classes. And it contains more than 160,000 images in total and at least 200 images for each subordinate class. Accompanying the dataset, we also propose an effective framework called HybridNet to exploit the label hierarchy for FGVC. Specifically, multiple granularity features are first extracted by dealing with the hierarchical labels separately. And then they are fused through explicit operation, e.g., Compact Bilinear Pooling, to form a unified representation for the ultimate recognition. The experimental results on the novel VegFru, the public FGVC-Aircraft and CUB-200-2011 indicate that HybridNet achieves one of the top performance on these datasets. The dataset and code are available at https://github.com/hshustc/vegfru.

Rich and dense human labeled datasets are the main enabling factor, among others, for the recent exciting work on vision-language understanding. Many seemingly distinct annotations (e.g., semantic segmentation and visual questions answering (VQA)) are inherently connected in that they reveal different levels and perspectives of human understandings about the same visual scenes --- and even the same set of MS COCO images. The popularity of MS COCO could strongly correlate those annotations and tasks. Explicitly linking them up, as we envision, can significantly benefit not only individual tasks but also the overarching goal of unified vision-language understand. We present the preliminary work of linking the instance segmentations provided by MS COCO to the questions and answers (QA) in the VQA dataset. We call the collected links visual questions and segmentation answers (VQS). They transfer human supervision between the previously separate tasks, offer more effective leverage to existing problems, and also open the door for new tasks and richer models. We study two applications of the VQS data in this paper: supervised attention for VQA and a novel question-focused semantic segmentation task. For the former, we obtain state-of-the-art results on the VQA real multiple-choice task by simply augmenting multilayer perceptrons with some attention features that are learned by using the segmentation-QA links as explicit supervision. To put the latter in perspective, we study two plausible methods and an oracle upper bound.

In this paper, we propose a unified end-to-end trainable multi-task network that jointly handles lane and road marking detection and recognition that is guided by a vanishing point under adverse weather conditions. We tackle rainy and low illumination conditions, which have not been extensively studied until now due to clear challenges. For example, images taken under rainy days are subject to low illumination, while wet roads cause light reflection and distort the appearance of lane and road markings. At night, color distortion occurs under limited illumination. As a result, no benchmark dataset exists and only a few developed algorithms work under poor weather conditions. To address this shortcoming, we build up a lane and road marking benchmark which consists of about 20,000 images with 17 lane and road marking classes under four different scenarios: no rain, rain, heavy rain, and night. We train and evaluate several versions of the proposed multi-task network and validate the importance of each task. The resulting approach, VPGNet, can detect and classify lanes and road markings, and predict a vanishing point with a single forward pass. Experimental results show that our approach achieves high accuracy and robustness under various conditions in real-time (20 fps). The benchmark and the VPGNet model will be publicly available.

Understanding human activity and being able to explain it in detail surpasses mere action classification by far in both complexity and value. The challenge is thus to describe an activity on the basis of its most fundamental constituents, the individual postures and their distinctive transitions. Supervised learning of such a fine-grained representation based on elementary poses is very tedious and does not scale. Therefore, we propose a completely unsupervised deep learning procedure based solely on video sequences, which starts from scratch without requiring pre-trained networks, predefined body models, or keypoints. A combinatorial sequence matching algorithm proposes relations between frames from subsets of the training data, while a CNN is reconciling the transitivity conflicts of the different subsets to learn a single concerted pose embedding despite changes in appearance across sequences. Without any manual annotation, the model learns a structured representation of postures and their temporal development. The model not only enables retrieval of similar postures but also temporal super-resolution. Additionally, based on a recurrent formulation, next frames can be synthesized.

We present an unsupervised representation learning approach using videos without semantic labels. We leverage the temporal coherence as a supervisory signal by formulating representation learning as a sequence sorting task. We take temporally shuffled frames (i.e. in non-chronological order) as inputs and train a convolutional neural network to sort the shuffled sequences. Similar to comparison-based sorting algorithms, we propose to extract features from all frame pairs and aggregate them to predict the correct order. As sorting shuffled image sequence requires an understanding of the statistical temporal structure of images, training with such a proxy task allows us to learn rich and generalizable visual representation. We validate the effectiveness of the learned representation using our method as pre-training on high-level recognition problems. The experimental results show that our method compares favorably against state-of-the-art methods on action recognition, image classification and object detection tasks.

We address an essential problem in computer vision, that of unsupervised foreground object segmentation in video, where a main object of interest in a video sequence should be automatically separated from its background. An efficient solution to this task would enable large-scale video interpretation at a high semantic level in the absence of the costly manual labeling. We propose an efficient unsupervised method for generating foreground object soft masks based on automatic selection and learning from highly probable positive features. We show that such features can be selected efficiently by taking into consideration the spatio-temporal appearance and motion consistency of the object in the video sequence. We also emphasize the role of the contrasting properties between the foreground object and its background. Our model is created over several stages: we start from pixel level analysis and move to descriptors that consider information over groups of pixels combined with efficient motion analysis. We also prove theoretical properties of our unsupervised learning method, which under some mild constraints is guaranteed to learn the correct classifier even in the unsupervised case. We achieve competitive and even state of the art results on the challenging Youtube-Objects and SegTrack datasets, while being at least one order of magnitude faster than the competition. We believe that the strong performance of our method, along with its theoretical properties, constitute a solid step towards solving unsupervised discovery in video.

In recent years, convolutional neural networks have shown its strong power for stereo matching cost learning. Current approaches learn the parameters of their models from public datasets with ground truth disparity. However, due to the limitations of these datasets and the difficulty of collecting new stereo data, current methods fail in real-life cases. In this work, we present a framework for learning stereo matching cost without human supervision. Our method updates the network parameter in a iterative manner. It starts with randomly initialized network. Correct matchings are carefully picked and used as training data in each round. In the end, the networks converges to a stable state, which performs comparably with supervised methods on various benchmarks.

A first-person camera, placed at a person's head, captures, which objects are important to the camera wearer. Most prior methods for this task learn to detect such important objects from the manually labeled first-person data in a supervised fashion. However, important objects are strongly related to the camera wearer's internal state such as his intentions and attention, and thus, only the person wearing the camera can provide the importance labels. Such a constraint makes the annotation process costly and limited in scalability. In this work, we show that we can detect important objects in first-person images without the supervision by the camera wearer or even third-person labelers. We formulate an important detection problem as an interplay between the 1) segmentation and 2) recognition agents. The segmentation agent first proposes a possible important object segmentation mask for each image, and then feeds it to the recognition agent, which learns to predict an important object mask using visual semantics and spatial features. We implement such an interplay between both agents via an alternating cross-pathway supervision scheme inside our proposed Visual-Spatial Network (VSN). Our VSN consists of spatial ("where") and visual ("what") pathways, one of which learns common visual semantics while the other focuses on the spatial location cues. Our unsupervised learning is accomplished via a cross-pathway supervision, where one pathway feeds its predictions to a segmentation agent, which proposes a candidate important object segmentation mask that is then used by the other pathway as a supervisory signal. We show our method's success on two different important object datasets, where our method achieves similar or better results as the supervised methods.

Unsupervised learning from visual data is one of the most difficult challenges in computer vision. It is essential for understanding how visual recognition works. Learning from unsupervised input has an immense practical value, as huge quantities of unlabeled videos can be collected at low cost. Here we address the task of unsupervised learning to detect and segment foreground objects in single images. We achieve our goal by training a student pathway, consisting of a deep neural network that learns to predict, from a single input image, the output of a teacher pathway that performs unsupervised object discovery in video. Our approach is different from the published methods that perform unsupervised discovery in videos or in collections of images at test time. We move the unsupervised discovery phase during the training stage, while at test time we apply the standard feed-forward processing along the student pathway. This has a dual benefit: firstly, it allows, in principle, unlimited generalization possibilities during training, while remaining fast at testing. Secondly, the student not only becomes able to detect in single images significantly better than its unsupervised video discovery teacher, but it also achieves state of the art results on two current benchmarks, YouTube Objects and Object Discovery datasets. At test time, our system is two orders of magnitude faster than other previous methods.

Despite rapid advances in face recognition, there remains a clear gap between the performance of still image-based face recognition and video-based face recognition, due to the vast difference in visual quality between the domains and the difficulty of curating diverse large-scale video datasets. This paper addresses both of those challenges, through an image to video feature-level domain adaptation approach, to learn discriminative video frame representations. The framework utilizes large-scale unlabeled video data to reduce the gap between different domains while transferring discriminative knowledge from large-scale labeled still images. Given a face recognition network that is pretrained in the image domain, the adaptation is achieved by (i) distilling knowledge from the network to a video adaptation network through feature matching, (ii) performing feature restoration through synthetic data augmentation and (iii) learning a domain-invariant feature through a domain adversarial discriminator. We further improve performance through a discriminator-guided feature fusion that boosts high-quality frames while eliminating those degraded by video domain-specific factors. Experiments on the YouTube Faces and IJB-A datasets demonstrate that each module contributes to our feature-level domain adaptation framework and substantially improves video face recognition performance to achieve state-of-the-art accuracy. We demonstrate qualitatively that the network learns to suppress diverse artifacts in videos such as pose, illumination or occlusion without being explicitly trained for them.

Recent ground-breaking works have shown that deep neural networks can be trained end-to-end to regress dense disparity maps directly from image pairs. Computer generated imagery is deployed to gather the large data corpus required to train such networks, an additional fine-tuning allowing to adapt the model to work well also on real and possibly diverse environments. Yet, besides a few public datasets such as Kitti, the ground-truth needed to adapt the network to a new scenario is hardly available in practice. In this paper we propose a novel unsupervised adaptation approach that enables to fine-tune a deep learning stereo model without any ground-truth information. We rely on off-the-shelf stereo algorithms together with state-of-the-art confidence measures, the latter able to ascertain upon correctness of the measurements yielded by former. Thus, we train the network based on a novel loss-function that penalizes predictions disagreeing with the highly confident disparities provided by the algorithm and enforces a smoothness constraint. Experiments on popular datasets (KITTI 2012, KITTI 2015 and Middlebury 2014) and other challenging test images demonstrate the effectiveness of our proposal.

This paper is the first to address the problem of unsupervised action localization in videos. Given unlabeled data without bounding box annotations, we propose a novel approach that: 1) Discovers action class labels and 2) Spatio-temporally localizes actions in videos. It begins by computing local video features to apply spectral clustering on a set of unlabeled training videos. For each cluster of videos, an undirected graph is constructed to extract a dominant set, which are known for high internal homogeneity and inhomogeneity between vertices outside it. Next, a discriminative clustering approach is applied, by training a classifier for each cluster, to iteratively select videos from the non dominant set and obtain complete video action classes. Once classes are discovered, training videos within each cluster are selected to perform automatic spatio-temporal annotations, by first oversegmenting videos in each discovered class into supervoxels and constructing a directed graph to apply a variant of knapsack problem with temporal constraints. Knapsack optimization jointly collects a subset of supervoxels, by enforcing the annotated action to be spatio-temporally connected and its volume to be the size of an actor. These annotations are used to train SVM action classifiers. During testing, actions are localized using a similar Knapsack approach, where supervoxels are grouped together and SVM, learned using videos from discovered action classes, is used to recognize these actions. We evaluate our approach on UCF Sports, Sub-JHMDB, JHMDB, THUMOS13 and UCF101 datasets. Our experiments suggest that despite using no action class labels and no bounding box annotations, we are able to get competitive results to the state-of-the-art supervised methods.

It has been recently shown that neural networks can recover the geometric structure of a face from a single given image. A common denominator of most existing face geometry reconstruction methods is the restriction of the solution space to some low-dimensional subspace. While such a model significantly simplifies the reconstruction problem, it is inherently limited in its expressiveness. As an alternative, we propose an Image-to-Image translation network that jointly maps the input image to a depth image and a facial correspondence map. This explicit pixel-based mapping can then be utilized to provide high quality reconstructions of diverse faces under extreme expressions, using a purely geometric refinement process. In the spirit of recent approaches, the network is trained only with synthetic data, and is then evaluated on in-the-wild facial images. Both qualitative and quantitative analyses demonstrate the accuracy and the robustness of our approach.

We propose a novel framework for abnormal event detection in video that requires no training sequences. Our framework is based on unmasking, a technique previously used for authorship verification in text documents, which we adapt to our task. We iteratively train a binary classifier to distinguish between two consecutive video sequences while removing at each step the most discriminant features. Higher training accuracy rates of the intermediately obtained classifiers represent abnormal events. To the best of our knowledge, this is the first work to apply unmasking for a computer vision task. We compare our method with several state-of-the-art supervised and unsupervised methods on four benchmark data sets. The empirical results indicate that our abnormal event detection framework can achieve state-of-the-art results, while running in real-time at 20 frames per second.

While deep learning is remarkably successful on perceptual tasks, it was also shown to be vulnerable to adversarial perturbations of the input. These perturbations denote noise added to the input that was generated specifically to fool the system while being quasi-imperceptible for humans. More severely, there even exist universal perturbations that are input-agnostic but fool the network on the majority of inputs. While recent work has focused on image classification, this work proposes attacks against semantic image segmentation: we present an approach for generating (universal) adversarial perturbations that make the network yield a desired target segmentation as output. We show empirically that there exist barely perceptible universal noise patterns which result in nearly the same predicted segmentation for arbitrary inputs. Furthermore, we also show the existence of universal noise which removes a target class (e.g., all pedestrians) from the segmentation while leaving the segmentation mostly unchanged otherwise.

This work addresses the problem of domain adaptation and generalization in a unified fashion. The main idea is to exploit the siamese architecture with the Contrastive Loss to address the domain shift and generalization problems. The framework is general, and can be used with any architecture. One of the main strengths of the approach is the "speed" of adaptation, which requires an extremely low number of labeled training samples from the target domain, even only one per category. The same architecture and loss function can be easily extended to domain generalization. We present state-of-the-art results for both of these applications.

While natural beauty is often considered a subjective property of images, in this paper, we take an objective approach and provide methods for quantifying and predicting the scenicness of an image. Using a dataset containing hundreds of thousands of outdoor images captured throughout Great Britain with crowdsourced ratings of natural beauty, we propose an approach to predict scenicness which explicitly accounts for the variance of human ratings. We demonstrate that quantitative measures of scenicness can benefit semantic image understanding, content-aware image processing, and a novel application of cross-view mapping, where the sparsity of ground-level images can be addressed by incorporating unlabeled overhead images in the training and prediction steps. For each application, our methods for scenicness prediction result in quantitative and qualitative improvements over baseline approaches.

Understanding where people look in images is an important problem in computer vision. Despite significant research, it remains unclear to what extent human fixations can be predicted by low-level (contrast) compared to high-level (presence of objects) image features. Here we address this problem by introducing two novel models that use different feature spaces but the same readout architecture. The first model predicts human fixations based on deep neural network features trained on object recognition. This model sets a new state-of-the art in fixation prediction by achieving top performance in area under the curve metrics on the MIT300 hold-out benchmark (AUC = 88%, sAUC = 77%, NSS = 2.34). The second model uses purely low-level (isotropic contrast) features. This model achieves better performance than all models not using features pre-trained on object recognition, making it a strong baseline to assess the utility of high-level features. We then evaluate and visualize which fixations are better explained by low-level compared to high-level image features. Surprisingly we find that a substantial proportion of fixations are better explained by the simple low-level model than the state-of-the-art model. Comparing different features within the same powerful readout architecture allows us to better understand the relevance of low- versus high-level features in predicting fixation locations, while simultaneously achieving state-of-the-art saliency prediction.