ECCV 2018 论文列表

Visual attention has shown usefulness in image captioning, with the goal of enabling a caption model to selectively focus on regions of interest. Existing models typically rely on top-down language information and learn attention implicitly by optimizing the captioning objectives. While somewhat effective, the learned top-down attention can fail to focus on correct regions of interest without direct supervision of attention. Inspired by the human visual system which is driven by not only the task-specific top-down signals but also the visual stimuli, we in this work propose to use both types of attention for image captioning. In particular, we highlight the complementary nature of the two types of attention and develop a model (Boosted Attention) to integrate them for image captioning. We validate the proposed approach with state-of-the-art performance across various evaluation metrics.

Human shape estimation is an important task for video editing, animation and fashion industry. Predicting 3D human body shape from natural images, however, is highly challenging due to factors such as variation in human bodies, clothing and viewpoint. Prior methods addressing this problem typically attempt to fit parametric body models with certain priors on pose and shape. In this work we argue for an alternative representation and propose BodyNet, a neural network for direct inference of volumetric body shape from a single image. BodyNet is an end-to-end trainable network that benefits from (i) a volumetric 3D loss, (ii) a multi-view re-projection loss, and (iii) intermediate supervision of 2D pose, 2D body part segmentation, and 3D pose. Each of them results in performance improvement as demonstrated by our experiments. To evaluate the method, we fit the SMPL model to our network output and show state-of-the-art results on the SURREAL and Unite the People datasets, outperforming recent approaches. Besides achieving state-of-the-art performance, our method also enables volumetric body-part segmentation.

This paper presents a network to detect shadows by exploring and combining global context in deep layers and local context in shallow layers of a deep convolutional neural network (CNN). There are two technical contributions in our network design. First, we formulate the recurrent attention residual (RAR) module to combine the contexts in two adjacent CNN layers and learn an attention map to select a residual and then refine the context features. Second, we develop a bidirectional feature pyramid network (BFPN) to aggregate shadow contexts spanned across different CNN layers by deploying two series of RAR modules in the network to iteratively combine and refine context features: one series to refine context features from deep to shallow layers, and another series from shallow to deep layers. Hence, we can better suppress false detections and enhance shadow details at the same time. We evaluate our network on two common shadow detection benchmark datasets: SBU and UCF. Experimental results show that our network outperforms the best existing method with 34.88% reduction on SBU and 34.57% reduction on UCF for the balance error rate.

Semantic segmentation requires both rich spatial information and sizeable receptive field. However, modern approaches usually compromise spatial resolution to achieve real-time inference speed, which leads to poor performance. In this paper, we address this dilemma with a novel Bilateral Segmentation Network (BiSeNet). We first design a Spatial Path with a small stride to preserve the spatial information and generate high-resolution features. Meanwhile, a Context Path with a fast downsampling strategy is employed to obtain sufficient receptive field. On top of the two paths, we introduce a new Feature Fusion Module to combine features efficiently. The proposed architecture makes a right balance between the speed and segmentation performance on Cityscapes, CamVid, and COCO-Stuff datasets. Specifically, for a 2048x1024 input, we achieve 68.4% Mean IOU on the Cityscapes test dataset with speed of 105 FPS on one NVIDIA Titan XP card, which is significantly faster than the existing methods with comparable performance.

Occlusions present a great challenge for pedestrian detection in practical applications. In this paper, we propose a novel approach to simultaneous pedestrian detection and occlusion estimation by regressing two bounding boxes to localize the full body as well as the visible part of a pedestrian respectively. For this purpose, we learn a deep convolutional neural network (CNN) consisting of two branches, one for full body estimation and the other for visible part estimation. The two branches are treated differently during training such that they are learned to produce complementary outputs which can be further fused to improve detection performance. The full body estimation branch is trained to regress full body regions for positive pedestrian proposals, while the visible part estimation branch is trained to regress visible part regions for both positive and negative pedestrian proposals. The visible part region of a negative pedestrian proposal is forced to shrink to its center. In addition, we introduce a new criterion for selecting positive training examples, which contributes largely to heavily occluded pedestrian detection. We validate the effectiveness of the proposed bi-box regression approach on the Caltech and CityPersons datasets. Experimental results show that our approach achieves promising performance for detecting both non-occluded and occluded pedestrians, especially heavily occluded ones.

In this work, we study the 1-bit convolutional neural networks (CNNs), of which both the weights and activations are binary. While being efficient, the classification accuracy of the current 1-bit CNNs is much worse compared with their counterpart real-valued CNN models on the large-scale dataset, like ImageNet. To shrink the performance gap between the 1-bit and real-valued CNN models, we propose a novel model, dubbed Bi-Real net, which connects the real activations (after the 1-bit convolution and/or BatchNorm layer, before the sign function) to that of the consecutive block, through an identity shortcut. Consequently, compared to the standard 1-bit CNN, the representational capability of the Bi-Real net is significantly enhanced, only with a negligible additional cost on computation. Moreover, we develop a specific training algorithm including three technical novelties for 1-bit CNNs. First, we derive a tight approximation to the derivative of the non-differentiable sign function with respect to activation. Second, we propose a magnitude-aware gradient with respect to weight to update the weight parameter. Last, we pre-train the real-valued CNN model with a clip function, rather than the ReLU function, to provide a better initialization for Bi-Real net. Experiments on ImageNet show that the Bi-Real net with proposed training algorithm achieves 56.4% and 62.2% top-1 accuracy with 18 layers and 34 layers, respectively, and achieves up to 23.9X memory saving and 17.0X computational reduction.

Confidence measures for stereo gained popularity in recent years due to their improved capability to detect outliers and the increasing number of applications exploiting these cues. In this field, convolutional neural networks achieved top-performance compared to other known techniques in the literature by processing local information to tell disparity assignments from outliers. Despite this outstanding achievements, all approaches rely on clues extracted with small receptive fields thus ignoring most of the overall image content. Therefore, in this paper, we propose to exploit nearby and farther clues available from image and disparity domains to obtain a more accurate confidence estimation. While local information is very effective for detecting high frequency patterns, it lacks insights from farther regions in the scene. On the other hand, enlarging the receptive field allows to include clues from farther regions but produces smoother uncertainty estimation, not particularly accurate when dealing with high frequency patterns. For these reasons, we propose in this paper a multi-stage cascaded network to combine the best of the two worlds. Extensive experiments on three datasets using three popular stereo algorithms prove that the proposed framework outperforms state-of-the-art confidence estimation techniques.

Employing part-level features offers fine-grained information for pedestrian image description. A prerequisite of part discovery is that each part should be well located. Instead of using external resources like pose estimator, we consider content consistency within each part for precise part location. Specifically, we target at learning discriminative part-informed features for person retrieval and make two contributions. (i) A network named Part-based Convolutional Baseline (PCB). Given an image input, it outputs a convolutional descriptor consisting of several part-level features. With a uniform partition strategy, PCB achieves competitive results with the state-of-the-art methods, proving itself as a strong convolutional baseline for person retrieval. (ii) A refined part pooling (RPP) method. Uniform partition inevitably incurs outliers in each part, which are in fact more similar to other parts. RPP re-assigns these outliers to the parts they are closest to, resulting in refined parts with enhanced within-part consistency. Experiment confirms that RPP allows PCB to gain another round of performance boost. For instance, on the Market-1501 dataset, we achieve (77.4+4.2)% mAP and (92.3+1.5)% rank-1 accuracy, surpassing the state of the art by a large margin.

This paper addresses the semantic instance segmentation task in the open-set conditions, where input images can contain known and unknown object classes. The training process of existing semantic instance segmentation methods requires annotation masks for all object instances, which is expensive to acquire or even infeasible in some realistic scenarios, where the number of categories may increase boundlessly. In this paper, we present a novel open-set semantic instance segmentation approach capable of segmenting all known and unknown object classes in images, based on the output of an object detector trained on known object classes. We formulate the problem using a Bayesian framework, where the posterior distribution is approximated with a simulated annealing optimization equipped with an efficient image partition sampler. We show empirically that our method is competitive with state-of-the-art supervised methods on known classes, but also performs well on unknown classes when compared with unsupervised methods.

Temporal action proposal generation is an important yet challenging problem, since temporal proposals with rich action content are indispensable for analysing real-world videos with long duration and high proportion irrelevant content. This problem requires methods not only generating proposals with precise temporal boundaries, but also retrieving proposals to cover truth action instances with high recall and high overlap using relatively fewer proposals. To address these difficulties, we introduce an effective proposal generation method, named Boundary-Sensitive Network (BSN), which adopts "local to global" fashion. Locally, BSN first locates temporal boundaries with high probabilities, then directly combines these boundaries as proposals. Globally, with Boundary-Sensitive Proposal feature, BSN retrieves proposals by evaluating the confidence of whether a proposal contains an action within its region. We conduct experiments on two challenging datasets: ActivityNet-1.3 and THUMOS14, where BSN outperforms other state-of-the-art temporal action proposal generation methods with high recall and high temporal precision. Finally, further experiments demonstrate that by combining existing action classifiers, our method significantly improves the state-of-the-art temporal action detection performance.

We propose a benchmark for 6D pose estimation of a rigid object from a single RGB-D input image. The training data consists of a texture-mapped 3D object model or images of the object in known 6D poses. The benchmark comprises of: i) eight datasets in a unified format that cover different practical scenarios, including two new datasets focusing on varying lighting conditions, ii) an evaluation methodology with a pose-error function that deals with pose ambiguities, iii) a comprehensive evaluation of 15 diverse recent methods that captures the status quo of the field, and iv) an online evaluation system that is open for continuous submission of new results. The evaluation shows that methods based on point-pair features currently perform best, outperforming template matching methods, learning-based methods and methods based on 3D local features. The project website is available at bop.felk.cvut.cz.

Temporal Action Localization (TAL) in untrimmed video is important for many applications. But it is very expensive to annotate the segment-level ground truth (action class and temporal boundary). This raises the interest of addressing TAL with weak supervision, namely only video-level annotations are available during training). However, the state-of-the-art weakly-supervised TAL methods only focus on generating good Class Activation Sequence (CAS) over time but conduct simple thresholding on CAS to localize actions. In this paper, we first develop a novel weakly-supervised TAL framework called AutoLoc to directly predict the temporal boundary of each action instance. We propose a novel Outer-Inner-Contrastive (OIC) loss to automatically discover the needed segment-level supervision for training such a boundary predictor. Our method achieves dramatically improved performance: under the IoU threshold 0.5, our method improves mAP on THUMOS'14 from 13.7% to 21.2% and mAP on ActivityNet from 7.4% to 27.3%. It is also very encouraging to see that our weakly-supervised method achieves comparable results with some fully-supervised methods.

Deep learning based image-to-image translation methods aim at learning the joint distribution of the two domains and finding transformations between them. Despite recent GAN (Generative Adversarial Network) based methods have shown compelling visual results, they are prone to fail at preserving image-objects and maintaining translation consistency when faced with large and complex domain shifts, which reduces their practicality on tasks such as generating large-scale training data for different domains. To address this problem, we purpose a weakly supervised structure-aware image-to-image translation network, which is composed of encoders, generators, discriminators and parsing nets for the two domains, respectively, in a unified framework. The purposed network generates more visually plausible images of a different domain compared to the competing methods on different image-translation tasks. In addition, we quantitatively evaluate different methods by training Faster-RCNN and YOLO with datasets generated from the image-translation results and demonstrate significant improvement of the detection accuracies by using the proposed image-object preserving network.

The thud of a bouncing ball, the onset of speech as lips open -- when visual and audio events occur together, it suggests that there might be a common, underlying event that produced both signals. In this paper, we argue that the visual and audio components of a video signal should be modeled jointly using a fused multisensory representation. We propose to learn such a representation in a self-supervised way, by training a neural network to predict whether video frames and audio are temporally aligned. We use this learned representation for three applications: (a) sound source localization, i.e. visualizing the source of sound in a video; (b) audio-visual action recognition; and (c) on/off-screen audio source separation, e.g. removing the off-screen translator's voice from a foreign official's speech. Code, models, and video results are available on our webpage: http://andrewowens.com/multisensory.

In this paper, we introduce a novel problem of audio-visual event localization in unconstrained videos. We define an audio-visual event as an event that is both visible and audible in a video segment. We collect an Audio-Visual Event (AVE) dataset to systemically investigate three temporal localization tasks: supervised and weakly-supervised audio-visual event localization, and cross-modality localization. We develop an audio-guided visual attention mechanism to explore audio-visual correlations, propose a dual multimodal residual network (DMRN) to fuse information over the two modalities, and introduce an audio-visual distance learning network to handle the cross-modality localization. Our experiments support the following findings: joint modeling of auditory and visual modalities outperforms independent modeling, the learned attention can capture semantics of sounding objects, temporal alignment is important for audio-visual fusion, the proposed DMRN is effective in fusing audio-visual features, and strong correlations between the two modalities enable cross-modality localization.

We present a new approach to modeling visual attributes. Prior work casts attributes in a similar role as objects, learning a latent representation where properties (e.g., sliced) are recognized by classifiers much in the way objects (e.g., apple) are. However, this common approach fails to separate the attributes observed during training from the objects with which they are composed, making it ineffectual when encountering new attribute-object compositions. Instead, we propose to model attributes as operators. Our approach learns a semantic embedding that explicitly factors out attributes from their accompanying objects, and also benefits from novel regularizers expressing attribute operators’ effects (e.g., blunt should undo the effects of sharp). Not only does our approach align conceptually with the linguistic role of attributes as modifiers, but it also generalizes to recognize unseen compositions of objects and attributes, We validate our approach on two challenging datasets and demonstrate significant improvements over the state-of-the-art. In addition, we show that not only can our model recognize unseen compositions robustly in an open-world setting, it can also generalize to compositions where objects themselves were unseen during training.

We are interested in attribute-guided face generation: given a low-res face input image, an attribute vector that can be extracted from a high-res image (attribute image), our new method generates a high-res face image for the low-res input that satisfies the given attributes. To address this problem, we condition the CycleGAN and propose conditional CycleGAN, which is designed to 1) handle unpaired training data because the training low/high-res and high-res attribute images may not necessarily align with each other, and to 2) allow easy control of the appearance of the generated face via the input attributes. We demonstrate impressive results on the attribute-guided conditional CycleGAN, which can synthesize realistic face images with appearance easily controlled by user-supplied attributes (e.g., gender, makeup, hair color, eyeglasses). Using the attribute image as identity to produce the corresponding conditional vector and by incorporating a face verification network, the attribute-guided network becomes the identity-guided conditional CycleGAN which produces impressive and interesting results on identity transfer. We demonstrate three applications on identity-guided conditional CycleGAN: identity-preserving face superresolution, face swapping, and frontal face generation, which consistently show the advantage of our new method.

Semantic correspondence is the problem of establishing correspondences across images depicting different instances of the same object or scene class. One of recent approaches to this problem is to estimate parameters of a global transformation model that densely aligns one image to the other. Since an entire correlation map between all feature pairs across images is typically used to predict such a global transformation, noisy features from different backgrounds, clutter, and occlusion distract the predictor from correct estimation of the alignment. This is a challenging issue, in particular, in the problem of semantic correspondence where a large degree of image variations is often involved. In this paper, we introduce an attentive semantic alignment method that focuses on reliable correlations, filtering out distractors. For effective attention, we also propose an offset-aware correlation kernel that learns to capture translation-invariant local transformations in computing correlation values over spatial locations. Experiments demonstrate the effectiveness of the attentive model and offset-aware kernel, and the proposed model combining both techniques achieves the state-of-the-art performance.

Recently, ensemble has been applied to deep metric learning to yield state-of-the-art results. Deep metric learning aims to learn deep neural networks for feature embeddings, distances of which satisfy given constraint. In deep metric learning, ensemble takes average of distances learned by multiple learners. As one important aspect of ensemble, the learners should be diverse in their feature embeddings. To this end, we propose an attention-based ensemble, which uses multiple attention masks, so that each learner can attend to different parts of the object. We also propose a divergence loss, which encourages diversity among the learners. The proposed method is applied to the standard benchmarks of deep metric learning and experimental results show that it outperforms the state-of-the-art methods by a significant margin on image retrieval tasks.