CVPR 2017 论文列表

Convnets have enabled significant progress in pedestrian detection recently, but there are still open questions regard- ing suitable architectures and training data. We revisit CNN design and point out key adaptations, enabling plain Fas- terRCNN to obtain state-of-the-art results on the Caltech dataset. To achieve further improvement from more and better data, we introduce CityPersons, a new set of person annotations on top of the Cityscapes dataset. The di- versity of CityPersons allows us for the first time to train one single CNN model that generalizes well over mul- tiple benchmarks. Moreover, with additional training with CityPersons, we obtain top results using FasterRCNN on Caltech, improving especially for more difficult cases (heavy occlusion and small scale) and providing higher loc- alization quality.

The chest X-ray is one of the most commonly accessible radiological examinations for screening and diagnosis of many lung diseases. A tremendous number of X-ray imaging studies accompanied by radiological reports are accumulated and stored in many modern hospitals' Picture Archiving and Communication Systems (PACS). On the other side, it is still an open question how this type of hospital-size knowledge database containing invaluable imaging informatics (i.e., loosely labeled) can be used to facilitate the data-hungry deep learning paradigms in building truly large-scale high precision computer-aided diagnosis (CAD) systems. In this paper, we present a new chest X-ray database, namely "ChestX-ray8", which comprises 108,948 frontal-view X-ray images of 32,717 unique patients with the text-mined eight disease image labels (where each image can have multi-labels), from the associated radiological reports using natural language processing. Importantly, we demonstrate that these commonly occurring thoracic diseases can be detected and even spatially-located via a unified weakly-supervised multi-label image classification and disease localization framework, which is validated using our proposed dataset. Although the initial quantitative results are promising as reported, deep convolutional neural network based "reading chest X-rays" (i.e., recognizing and locating the common disease patterns trained with only image-level labels) remains a strenuous task for fully-automated high precision CAD systems.

Given the recent advances in depth prediction from Convolutional Neural Networks (CNNs), this paper investigates how predicted depth maps from a deep neural network can be deployed for the goal of accurate and dense monocular reconstruction. We propose a method where CNN-predicted dense depth maps are naturally fused together with depth measurements obtained from direct monocular SLAM, based on a scheme that privileges depth prediction in image locations where monocular SLAM approaches tend to fail, e.g. along low-textured regions, and vice-versa. We demonstrate the use of depth prediction to estimate the absolute scale of the reconstruction, hence overcoming one of the major limitations of monocular SLAM. Finally, we propose a framework to efficiently fuse semantic labels, obtained from a single frame, with dense SLAM, so to yield semantically coherent scene reconstruction from a single view. Evaluation results on two benchmark datasets show the robustness and accuracy of our approach.

Stereo matching is a well researched area using visible-band color cameras. Thermal images are typically lower resolution, have less texture, and are noisier compared to their visible-band counterparts and are more challenging for stereo matching algorithms. Previous benchmarks for stereo matching either focus entirely on visible-band cameras or contain only a single thermal camera. We present the Color And Thermal Stereo (CATS) benchmark, a dataset consisting of stereo thermal, stereo color, and cross-modality image pairs with high accuracy ground truth (< 2mm) generated from a LiDAR. We scanned 100 cluttered indoor and 80 outdoor scenes featuring challenging environments and conditions. CATS contains approximately 1400 images of pedestrians, vehicles, electronics, and other thermally interesting objects in different environmental conditions, including nighttime, daytime, and foggy scenes. Ground truth was projected to each of the four cameras to generate color-color, thermal-thermal, and cross-modality disparity maps. We develop a semi-automatic LiDAR to camera alignment procedure that does not require a calibration target. We compare state-of-the-art algorithms to baseline the dataset and show that in the thermal and cross modalities there is still much room for improvement. We expect our dataset to provide researchers with a more diverse set of imaged locations, objects, and modalities than previous benchmarks for stereo matching.

Deep neural networks require a large amount of labeled training data during supervised learning. However, collecting and labeling so much data might be infeasible in many cases. In this paper, we introduce a deep transfer learning scheme, called selective joint fine-tuning, for improving the performance of deep learning tasks with insufficient training data. In this scheme, a target learning task with insufficient training data is carried out simultaneously with another source learning task with abundant training data. However, the source learning task does not use all existing training data. Our core idea is to identify and use a subset of training images from the original source learning task whose low-level characteristics are similar to those from the target learning task, and jointly fine-tune shared convolutional layers for both tasks. Specifically, we compute descriptors from linear or nonlinear filter bank responses on training images from both tasks, and use such descriptors to search for a desired subset of training samples for the source learning task. Experiments demonstrate that our deep transfer learning scheme achieves state-of-the-art performance on multiple visual classification tasks with insufficient training data for deep learning. Such tasks include Caltech 256, MIT Indoor 67, and fine-grained classification problems (Oxford Flowers 102 and Stanford Dogs 120). In comparison to fine-tuning without a source domain, the proposed method can improve the classification accuracy by 2% - 10% using a single model. Codes and models are available at https://github.com/ZYYSzj/Selective-Joint-Fine-tuning.

In recent years, there has been a renewed interest in jointly modeling perception and action. At the core of this investigation is the idea of modeling affordances. However, when it comes to predicting affordances, even the state of the art approaches still do not use any ConvNets. Why is that? Unlike semantic or 3D tasks, there still does not exist any large-scale dataset for affordances. In this paper, we tackle the challenge of creating one of the biggest dataset for learning affordances. We use seven sitcoms to extract a diverse set of scenes and how actors interact with different objects in the scenes. Our dataset consists of more than 10K scenes and 28K ways humans can interact with these 10K images. We also propose a two-step approach to predict affordances in a new scene. In the first step, given a location in the scene we classify which of the 30 pose classes is the likely affordance pose. Given the pose class and the scene, we then use a Variational Autoencoder (VAE) to extract the scale and deformation of the pose. The VAE allows us to sample the distribution of possible poses at test time. Finally, we show the importance of large-scale data in learning a generalizable and robust model of affordances.

Despite the recent success of neural networks in image feature learning, a major problem in the video domain is the lack of sufficient labeled data for learning to model temporal information. In this paper, we propose an unsupervised temporal modeling method that learns from untrimmed videos. The speed of motion varies constantly, e.g., a man may run quickly or slowly. We therefore train a Multirate Visual Recurrent Model (MVRM) by encoding frames of a clip with different intervals. This learning process makes the learned model more capable of dealing with motion speed variance. Given a clip sampled from a video, we use its past and future neighboring clips as the temporal context, and reconstruct the two temporal transitions, i.e., present->past transition and present->future transition, reflecting the temporal information in different views. The proposed method exploits the two transitions simultaneously by incorporating a bidirectional reconstruction which consists of a backward reconstruction and a forward reconstruction. We apply the proposed method to two challenging video tasks, i.e., complex event detection and video captioning, in which it achieves state-of-the-art performance. Notably, our method generates the best single feature for event detection with a relative improvement of 10.4% on the MEDTest-13 dataset and achieves the best performance in video captioning across all evaluation metrics on the YouTube2Text dataset.

Person re-identification (ReID) is an important task in wide area video surveillance which focuses on identifying people across different cameras. Recently, deep learning networks with a triplet loss become a common framework for person ReID. However, the triplet loss pays main attentions on obtaining correct orders on the training set. It still suffers from a weaker generalization capability from the training set to the testing set, thus resulting in inferior performance. In this paper, we design a quadruplet loss, which can lead to the model output with a larger inter-class variation and a smaller intra-class variation compared to the triplet loss. As a result, our model has a better generalization ability and can achieve a higher performance on the testing set. In particular, a quadruplet deep network using a margin-based online hard negative mining is proposed based on the quadruplet loss for the person ReID. In extensive experiments, the proposed network outperforms most of the state-of-the-art algorithms on representative datasets which clearly demonstrates the effectiveness of our proposed method.

Among learning based hashing methods, supervised hashing seeks compact binary representation of the training data to preserve semantic similarities. Recent years have witnessed various problem formulations and optimization methods for supervised hashing. Most of them optimize a form of loss function with a regulization term, which can be viewed as a maximum a posterior (MAP) estimation of the hashing codes. However, these approaches are prone to overfitting unless hyperparameters are tuned carefully. To address this problem, we present a novel fully Bayesian treatment for supervised hashing problem, named Bayesian Supervised Hashing (BSH), in which hyperparameters are automatically tuned during optimization. Additionally, by utilizing automatic relevance determination (ARD), we can figure out relative discriminating ability of different hashing bits and select most informative bits among them. Experimental results on three real-world image datasets with semantic information show that BSH can achieve superior performance over state-of-the-art methods with comparable training time.

There is more to images than their objective physical content: for example, advertisements are created to persuade a viewer to take a certain action. We propose the novel problem of automatic advertisement understanding. To enable research on this problem, we create two datasets: an image dataset of 64,832 image ads, and a video dataset of 3,477 ads. Our data contains rich annotations encompassing the topic and sentiment of the ads, questions and answers describing what actions the viewer is prompted to take and the reasoning that the ad presents to persuade the viewer ("What should I do according to this ad, and why should I do it?"), and symbolic references ads make (e.g. a dove symbolizes peace). We also analyze the most common persuasive strategies ads use, and the capabilities that computer vision systems should have to understand these strategies. We present baseline classification results for several prediction tasks, including automatically answering questions about the messages of the ads.

We present a novel visual attention tracking technique based on Shared Attention modeling. By considering the viewer as a participant in the activity occurring in the scene, our model learns the loci of attention of the scene actors and use it to augment image salience. We go beyond image salience and instead of only computing the power of image regions to pull attention, we also consider the strength with which the scene actors push attention to the region in question, thus the term Attentional Push. We present a convolutional neural network (CNN) which augments standard saliency models with Attentional Push. Our model contains two pathways: an Attentional Push pathway which learns the gaze location of the scene actors and a saliency pathway. These are followed by a shallow augmented saliency CNN which combines them and generates the augmented saliency. For training, we use transfer learning to initialize and train the Attentional Push CNN by minimizing the classification error of following the actors' gaze location on a 2-D grid using a large-scale gaze-following dataset. The Attentional Push CNN is then fine-tuned along with the augmented saliency CNN to minimize the Euclidean distance between the augmented saliency and ground truth fixations using an eye-tracking dataset, annotated with the head and the gaze location of the scene actors. We evaluate our model on three challenging eye fixation datasets, SALICON, iSUN and CAT2000, and illustrate significant improvements in predicting viewers' fixations in social scenes.

We address personalization issues of image captioning, which have not been discussed yet in previous research. For a query image, we aim to generate a descriptive sentence, accounting for prior knowledge such as the user's active vocabularies in previous documents. As applications of personalized image captioning, we tackle two post automation tasks: hashtag prediction and post generation, on our newly collected Instagram dataset, consisting of 1.1M posts from 6.3K users. We propose a novel captioning model named Context Sequence Memory Network (CSMN). Its unique updates over previous memory network models include (i) exploiting memory as a repository for multiple types of context information, (ii) appending previously generated words into memory to capture long-term information without suffering from the vanishing gradient problem, and (iii) adopting CNN memory structure to jointly represent nearby ordered memory slots for better context understanding. With quantitative evaluation and user studies via Amazon Mechanical Turk, we show the effectiveness of the three novel features of CSMN and its performance enhancement for personalized image captioning over state-of-the-art captioning models.

We introduce the task of Multi-Modal Machine Comprehension (M3C), which aims at answering multimodal questions given a context of text, diagrams and images. We present the Textbook Question Answering (TQA) dataset that includes 1,076 lessons and 26,260 multi-modal questions, taken from middle school science curricula. Our analysis shows that a significant portion of questions require complex parsing of the text and the diagrams and reasoning, indicating that our dataset is more complex compared to previous machine comprehension and visual question answering datasets. We extend state-of-the-art methods for textual machine comprehension and visual question answering to the TQA dataset. Our experiments show that these models do not perform well on TQA. The presented dataset opens new challenges for research in question answering and reasoning across multiple modalities.

For survival, a living agent (e.g., human in Fig. 1(a)) must have the ability to assess risk (1) by temporally anticipating accidents before they occur (Fig. 1(b)), and (2) by spatially localizing risky regions (Fig. 1(c)) in the environment to move away from threats. In this paper, we take an agent-centric approach to study the accident anticipation and risky region localization tasks. We propose a novel soft-attention Recurrent Neural Network (RNN) which explicitly models both spatial and appearance-wise non-linear interaction between the agent triggering the event and another agent or static-region involved. In order to test our proposed method, we introduce the Epic Fail (EF) dataset consisting of 3000 viral videos capturing various accidents. In the experiments, we evaluate the risk assessment accuracy both in the temporal domain (accident anticipation) and spatial domain (risky region localization) on our EF dataset and the Street Accident (SA) dataset. Our method consistently outperforms other baselines on both datasets.

"If I provide you a face image of mine (without telling you the actual age when I took the picture) and a large amount of face images that I crawled (containing labeled faces of different ages but not necessarily paired), can you show me what I would look like when I am 80 or what I was like when I was 5?" The answer is probably a "No." Most existing face aging works attempt to learn the transformation between age groups and thus would require the paired samples as well as the labeled query image. In this paper, we look at the problem from a generative modeling perspective such that no paired samples is required. In addition, given an unlabeled image, the generative model can directly produce the image with desired age attribute. We propose a conditional adversarial autoencoder (CAAE) that learns a face manifold, traversing on which smooth age progression and regression can be realized simultaneously. In CAAE, the face is first mapped to a latent vector through a convolutional encoder, and then the vector is projected to the face manifold conditional on age through a deconvolutional generator. The latent vector preserves personalized face features (i.e., personality) and the age condition controls progression vs. regression. Two adversarial networks are imposed on the encoder and generator, respectively, forcing to generate more photo-realistic faces. Experimental results demonstrate the appealing performance and flexibility of the proposed framework by comparing with the state-of-the-art and ground truth.

Generalization performance of trained computer vision (CV) systems that use computer graphics (CG) generated data is not yet effective due to the concept of 'domain-shift' between virtual and real data. Although simulated data augmented with a few real-world samples has been shown to mitigate domain shift and improve transferability of trained models, guiding or bootstrapping the virtual data generation with the distributions learnt from target real world domain is desired, especially in the fields where annotating even few real images is laborious (such as semantic labeling, optical flow, and intrinsic images etc.). In order to address this problem in an unsupervised manner, our work combines recent advances in CG, which aims at generating stochastic scene layouts using large collections of 3D object models, and generative adversarial training, which aims at training generative models by measuring discrepancy between generated and real data in terms of their separability in the space of a deep discriminatively-trained classifier. Our method uses iterative estimation of the posterior density of prior distributions for a generative graphical model. This is done within a rejection sampling framework. Initially, we assume uniform distributions as priors over parameters of a scene described by a generative graphical model. As iterations proceed the uniform prior distributions are updated sequentially to distributions that are closer to the unknown distributions of target data. We demonstrate the utility of adversarially tuned scene generation on two real world benchmark datasets (CityScapes and CamVid) for traffic scene semantic labeling with a deep convolutional net (DeepLab). We obtained performance improvements by 2.28 and 3.14 points on the IoU metric between the DeepLab models trained on simulated sets prepared from the scene generation models before and after tuning to CityScapes and CamVid respectively.

This paper proposes a novel tracker which is controlled by sequentially pursuing actions learned by deep reinforcement learning. In contrast to the existing trackers using deep networks, the proposed tracker is designed to achieve a light computation as well as satisfactory tracking accuracy in both location and scale. The deep network to control actions is pre-trained using various training sequences and fine-tuned during tracking for online adaptation to target and background changes. The pre-training is done by utilizing deep reinforcement learning as well as supervised learning. The use of reinforcement learning enables even partially labeled data to be successfully utilized for semi-supervised learning. Through evaluation of the OTB dataset, the proposed tracker is validated to achieve a competitive performance that is three times faster than state-of-the-art, deep network-based trackers. The fast version of the proposed method, which operates in real-time on GPU, outperforms the state-of-the-art real-time trackers.

Calibrating the intrinsic properties of a camera is one of the fundamental tasks required for a variety of computer vision and image processing tasks. The precise measurement of focal length, location of the principal point as well as distortion parameters of the lens is crucial, for example, for 3D reconstruction. Although a variety of methods exist to achieve this goal, they are often cumbersome to carry out, require substantial manual interaction, expert knowledge, and a significant operating volume. We propose a novel calibration method based on the usage of directionally encoded light rays for estimating the intrinsic parameters. It enables a fully automatic calibration with a small device mounted close to the front lens element and still enables an accuracy comparable to standard methods even when the lens is focused up to infinity. Our method overcomes the mentioned limitations since it guarantees an accurate calibration without any human intervention while requiring only a limited amount of space. Besides that, the approach also allows to estimate the distance of the focal plane as well as the size of the aperture. We demonstrate the advantages of the proposed method by evaluating several camera/lens configurations using prototypical devices.

Accurate estimation of camera matrices is an important step in structure from motion algorithms. In this paper we introduce a novel rank constraint on collections of fundamental matrices in multi-view settings. We show that in general, with the selection of proper scale factors, a matrix formed by stacking fundamental matrices between pairs of images has rank 6. Moreover, this matrix forms the symmetric part of a rank 3 matrix whose factors relate directly to the corresponding camera matrices. We use this new characterization to produce better estimations of fundamental matrices by optimizing an L1-cost function using Iterative Re-weighted Least Squares and Alternate Direction Method of Multiplier. We further show that this procedure can improve the recovery of camera locations, particularly in multi-view settings in which fewer images are available.