CVPR 2021 论文列表

In this paper we propose a new problem scenario in image processing, wide-range image blending, which aims to smoothly merge two different input photos into a panorama by generating novel image content for the intermediate region between them. Although such problem is closely related to the topics of image inpainting, image outpainting, and image blending, none of the approaches from these topics is able to easily address it. We introduce an effective deep-learning model to realize wide-range image blending, where a novel Bidirectional Content Transfer module is proposed to perform the conditional prediction for the feature representation of the intermediate region via recurrent neural networks. In addition to ensuring the spatial and semantic consistency during the blending, we also adopt the contextual attention mechanism as well as the adversarial learning scheme in our proposed method for improving the visual quality of the resultant panorama. We experimentally demonstrate that our proposed method is not only able to produce visually appealing results for wide-range image blending, but also able to provide superior performance with respect to several baselines built upon the state-of-the-art image inpainting and outpainting approaches.

This paper presents a novel method, termed Bridge to Answer, to infer correct answers for questions about a given video by leveraging adequate graph interactions of heterogeneous crossmodal graphs. To realize this, we learn question conditioned visual graphs by exploiting the relation between video and question to enable each visual node using question-to-visual interactions to encompass both visual and linguistic cues. In addition, we propose bridged visual-to-visual interactions to incorporate two complementary visual information on appearance and motion by placing the question graph as an intermediate bridge. This bridged architecture allows reliable message passing through compositional semantics of the question to generate an appropriate answer. As a result, our method can learn the question conditioned visual representations attributed to appearance and motion that show powerful capability for video question answering. Extensive experiments prove that the proposed method provides effective and superior performance than state-of-the-art methods on several benchmarks.

Recent advances in neuroscience have highlighted the effectiveness of multi-modal medical data for investigating certain pathologies and understanding human cognition. However, obtaining full sets of different modalities is limited by various factors, such as long acquisition times, high examination costs and artifact suppression. In addition, the complexity, high dimensionality and heterogeneity of neuroimaging data remains another key challenge in leveraging existing randomized scans effectively, as data of the same modality is often measured differently by different machines. There is a clear need to go beyond the traditional imaging-dependent process and synthesize anatomically specific target-modality data from a source input. In this paper, we propose to learn dedicated features that cross both intre- and intra-modal variations using a novel CSCl_4Net. Through an initial unification of intra-modal data in the feature maps and multivariate canonical adaptation, CSCl_4Net facilitates feature-level mutual transformation. The positive definite Riemannian manifold-penalized data fidelity term further enables CSCl_4Net to reconstruct missing measurements according to transformed features. Finally, the maximization l_4-norm boils down to a computationally efficient optimization problem. Extensive experiments validate the ability and robustness of our CSCl_4Net compared to the state-of-the-art methods on multiple datasets.

We present a high-performance method that can achieve mask-level instance segmentation with only bounding-box annotations for training. While this setting has been studied in the literature, here we show significantly stronger performance with a simple design (e.g., dramatically improving previous best reported mask AP of 21.1% to 31.6% on the COCO dataset). Our core idea is to redesign the loss of learning masks in instance segmentation, with no modification to the segmentation network itself. The new loss functions can supervise the mask training without relying on mask annotations. This is made possible with two loss terms, namely, 1) a surrogate term that minimizes the discrepancy between the projections of the ground-truth box and the predicted mask; 2) a pairwise loss that can exploit the prior that proximal pixels with similar colors are very likely to have the same category label. Experiments demonstrate that the redesigned mask loss can yield surprisingly high-quality instance masks with only box annotations. For example, without using any mask annotations, with a ResNet-101 backbone and 3x training schedule, we achieve 33.2% mask AP on COCO test-dev split (vs. 39.1% of the fully supervised counterpart). Our excellent experiment results on COCO and Pascal VOC indicate that our method dramatically narrows the performance gap between weakly and fully supervised instance segmentation. Code is available at https://git.io/AdelaiDet

We present Boundary IoU (Intersection-over-Union), a new segmentation evaluation measure focused on boundary quality. We perform an extensive analysis across different error types and object sizes and show that Boundary IoU is significantly more sensitive than the standard Mask IoU measure to boundary errors for large objects and does not over-penalize errors on smaller objects. The new quality measure displays several desirable characteristics like symmetry w.r.t. prediction/ground truth pairs and balanced responsiveness across scales, which makes it more suitable for segmentation evaluation than other boundary-focused measures like Trimap IoU and F-measure. Based on Boundary IoU, we update the standard evaluation protocols for instance and panoptic segmentation tasks by proposing the Boundary AP (Average Precision) and Boundary PQ (Panoptic Quality) metrics, respectively. Our experiments show that the new evaluation metrics track boundary quality improvements that are generally overlooked by current Mask IoU-based evaluation metrics. We hope that the adoption of the new boundary-sensitive evaluation metrics will lead to rapid progress in segmentation methods that improve boundary quality.

Referring image segmentation aims to segment the referent that is the corresponding object or stuff referred by a natural language expression in an image. Its main challenge lies in how to effectively and efficiently differentiate between the referent and other objects of the same category as the referent. In this paper, we tackle the challenge by jointly performing compositional visual reasoning and accurate segmentation in a single stage via the proposed novel Bottom-Up Shift (BUS) and Bidirectional Attentive Refinement (BIAR) modules. Specifically, BUS progressively locates the referent along hierarchical reasoning steps implied by the expression. At each step, it locates the corresponding visual region by disambiguating between similar regions, where the disambiguation bases on the relationships between regions. By the explainable visual reasoning, BUS explicitly aligns linguistic components with visual regions so that it can identify all the mentioned entities in the expression. BIAR fuses multi-level features via a two-way attentive message passing, which captures the visual details relevant to the referent to refine segmentation results. Experimental results demonstrate that the proposed method consisting of BUS and BIAR modules, can not only consistently surpass all existing state-of-the-art algorithms across common benchmark datasets but also visualize interpretable reasoning steps for stepwise segmentation. Code is available at https://github.com/incredibleXM/BUSNet.

In this paper, we are interested in the bottom-up paradigm of estimating human poses from an image. We study the dense keypoint regression framework that is previously inferior to the keypoint detection and grouping framework. Our motivation is that regressing keypoint positions accurately needs to learn representations that focus on the keypoint regions. We present a simple yet effective approach, named disentangled keypoint regression (DEKR). We adopt adaptive convolutions through pixel-wise spatial transformer to activate the pixels in the keypoint regions and accordingly learn representations from them. We use a multi-branch structure for separate regression: each branch learns a representation with dedicated adaptive convolutions and regresses one keypoint. The resulting disentangled representations are able to attend to the keypoint regions, respectively, and thus the keypoint regression is spatially more accurate. We empirically show that the proposed direct regression method outperforms keypoint detection and grouping methods and achieves superior bottom-up pose estimation results on two benchmark datasets, COCO and CrowdPose. The code and models are available at https://github.com/HRNet/DEKR.

We present BoTNet, a conceptually simple yet powerful backbone architecture that incorporates self-attention for multiple computer vision tasks including image classification, object detection and instance segmentation. By just replacing the spatial convolutions with global self-attention in the final three bottleneck blocks of a ResNet and no other changes, our approach improves upon the baselines significantly on instance segmentation and object detection while also reducing the parameters, with minimal overhead in latency. Through the design of BoTNet, we also point out how ResNet bottleneck blocks with self-attention can be viewed as Transformer blocks. Without any bells and whistles, BoTNet achieves 44.4% Mask AP and 49.7% Box AP on the COCO Instance Segmentation benchmark using the Mask R-CNN framework; surpassing the previous best published single model and single scale results of ResNeSt evaluated on the COCO validation set. Finally, we present a simple adaptation of the BoTNet design for image classification, resulting in models that achieve a strong performance of 84.7% top-1 accuracy on the ImageNet benchmark while being up to 1.64x faster in compute time than the popular EfficientNet models on TPU-v3 hardware. We hope our simple and effective approach will serve as a strong baseline for future research in self-attention models for vision.

Video content is multifaceted, consisting of objects, scenes, interactions or actions. The existing datasets mostly label only one of the facets for model training, resulting in the video representation that biases to only one facet depending on the training dataset. There is no study yet on how to learn a video representation from multifaceted labels, and whether multifaceted information is helpful for video representation learning. In this paper, we propose a new learning framework, MUlti-Faceted Integration (MUFI), to aggregate facets from different datasets for learning a representation that could reflect the full spectrum of video content. Technically, MUFI formulates the problem as visual-semantic embedding learning, which explicitly maps video representation into a rich semantic embedding space, and jointly optimizes video representation from two perspectives. One is to capitalize on the intra-facet supervision between each video and its own label descriptions, and the second predicts the "semantic representation" of each video from the facets of other datasets as the inter-facet supervision. Extensive experiments demonstrate that learning 3D CNN via our MUFI framework on a union of four large-scale video datasets plus two image datasets leads to superior capability of video representation. The pre-learnt 3D CNN with MUFI also shows clear improvements over other approaches on several downstream video applications. More remarkably, MUFI achieves 98.1%/80.9% on UCF101/HMDB51 for action recognition and 101.5% in terms of CIDEr-D score on MSVD for video captioning.

Neural networks have shown great abilities in estimating depth from a single image. However, the inferred depth maps are well below one-megapixel resolution and often lack fine-grained details, which limits their practicality. Our method builds on our analysis on how the input resolution and the scene structure affects depth estimation performance. We demonstrate that there is a trade-off between a consistent scene structure and the high-frequency details, and merge low- and high-resolution estimations to take advantage of this duality using a simple depth merging network. We present a double estimation method that improves the whole-image depth estimation and a patch selection method that adds local details to the final result. We demonstrate that by merging estimations at different resolutions with changing context, we can generate multi-megapixel depth maps with a high level of detail using a pre-trained model.

Neural network ensembles are gaining popularity by harnessing the complementary wisdom of multiple base models. Ensemble teams with high diversity promote high failure independence, which is effective for boosting the overall ensemble accuracy. This paper provides an in-depth study on how to design and compute ensemble diversity, which can capture the complementary decision capacity of ensemble member models. We make three original contributions. First, we revisit the ensemble diversity metrics in the literature and analyze the inherent problems of poor correlation between ensemble diversity and ensemble accuracy, which leads to the low quality ensemble selection using such diversity metrics. Second, instead of computing diversity scores for ensemble teams of different sizes using the same criteria, we introduce focal model based ensemble diversity metrics, coined as FQ-diversity metrics. Our new metrics significantly improve the intrinsic correlation between high ensemble diversity and high ensemble accuracy. Third, we introduce a diversity fusion method, coined as the EQ-diversity metric, by integrating the top three most representative FQ-diversity metrics. Comprehensive experiments on two benchmark datasets (CIFAR-10 and ImageNet) show that our FQ and EQ diversity metrics are effective for selecting high diversity ensemble teams to boost overall ensemble accuracy.

We propose a novel learned deep prior of body motion for 3D hand shape synthesis and estimation in the domain of conversational gestures. Our model builds upon the insight that body motion and hand gestures are strongly correlated in non-verbal communication settings. We formulate the learning of this prior as a prediction task of 3D hand shape over time given body motion input alone. Trained with 3D pose estimations obtained from a large-scale dataset of internet videos, our hand prediction model produces convincing 3D hand gestures given only the 3D motion of the speaker's arms as input. We demonstrate the efficacy of our method on hand gesture synthesis from body motion input, and as a strong body prior for single-view image-based 3D hand pose estimation. We demonstrate that our method outperforms previous state-of-the-art approaches and can generalize beyond the monologue-based training data to multi-person conversations.

We consider the challenging multi-person 3D body mesh estimation task in this work. Existing methods are mostly two-stage based--one stage for person localization and the other stage for individual body mesh estimation, leading to redundant pipelines with high computation cost and degraded performance for complex scenes (e.g., occluded person instances). In this work, we present a single stage model, Body Meshes as Points (BMP), to simplify the pipeline and lift both efficiency and performance. In particular, BMP adopts a new method that represents multiple person instances as points in the spatial-depth space where each point is associated with one body mesh. Hinging on such representations, BMP can directly predict body meshes for multiple persons in a single stage by concurrently localizing person instance points and estimating the corresponding body meshes. To better reason about depth ordering of all the persons within the same scene, BMP designs a simple yet effective inter-instance ordinal depth loss to obtain depth-coherent body mesh estimation. BMP also introduces a novel keypoint-aware augmentation to enhance model robustness to occluded person instances. Comprehensive experiments on benchmarks Panoptic, MuPoTS-3D and 3DPW clearly demonstrate the state-of-the-art efficiency of BMP for multi-person body mesh estimation, together with outstanding accuracy. Code can be found at: https://github.com/jfzhang95/BMP.

Generative adversarial networks (GANs) have gained considerable attention owing to their ability to reproduce images. However, they can recreate training images faithfully despite image degradation in the form of blur, noise, and compression, generating similarly degraded images. To solve this problem, the recently proposed noise robust GAN (NR-GAN) provides a partial solution by demonstrating the ability to learn a clean image generator directly from noisy images using a two-generator model comprising image and noise generators. However, its application is limited to noise, which is relatively easy to decompose owing to its additive and reversible characteristics, and its application to irreversible image degradation, in the form of blur, compression, and combination of all, remains a challenge. To address these problems, we propose blur, noise, and compression robust GAN (BNCR-GAN) that can learn a clean image generator directly from degraded images without knowledge of degradation parameters (e.g., blur kernel types, noise amounts, or quality factor values). Inspired by NR-GAN, BNCR-GAN uses a multiple-generator model composed of image, blur-kernel, noise, and quality-factor generators. However, in contrast to NR-GAN, to address irreversible characteristics, we introduce masking architectures adjusting degradation strength values in a data-driven manner using bypasses before and after degradation. Furthermore, to suppress uncertainty caused by the combination of blur, noise, and compression, we introduce adaptive consistency losses imposing consistency between irreversible degradation processes according to the degradation strengths. We demonstrate the effectiveness of BNCR-GAN through large-scale comparative studies on CIFAR-10 and a generality analysis on FFHQ. In addition, we demonstrate the applicability of BNCR-GAN in image restoration.

For several vision and robotics applications, 3D geometry of man-made environments such as indoor scenes can be represented with a small number of dominant planes. However, conventional 3D vision techniques typically first acquire dense 3D point clouds before estimating the compact piece-wise planar representations (e.g., by plane-fitting). This approach is costly, both in terms of acquisition and computational requirements, and potentially unreliable due to noisy point clouds. We propose Blocks-World Cameras, a class of imaging systems which directly recover dominant planes of piece-wise planar scenes (Blocks-World), without requiring point clouds. The Blocks-World Cameras are based on a structured-light system projecting a single pattern with a sparse set of cross-shaped features. We develop a novel geometric algorithm for recovering scene planes without explicit correspondence matching, thereby avoiding computationally intensive search or optimization routines. The proposed approach has low device and computational complexity, and requires capturing only one or two images. We demonstrate highly efficient and precise planar-scene sensing with simulations and real experiments, across various imaging conditions, including defocus blur, large lighting variations, ambient illumination, and scene clutter.

Blind deblurring has received considerable attention in recent years. However, state-of-the-art methods often fail to process saturated blurry images. The main reason is that saturated pixels are not conforming to the commonly used linear blur model. Pioneer arts suggest excluding saturated pixels during the deblurring process, which sacrifices the informative edges from saturated regions and results in insufficient information for kernel estimation when large saturated regions exist. To address this problem, we introduce a new blur model to fit both saturated and unsaturated pixels, and all informative pixels can be considered during deblurring process. Based on our model, we develop an effective maximum a posterior (MAP)-based optimization framework. Quantitative and qualitative evaluations on benchmark datasets and challenging real-world examples show that the proposed method performs favorably against existing methods.

We propose D-RISE, a method for generating visual explanations for the predictions of object detectors. Utilizing the proposed similarity metric that accounts for both localization and categorization aspects of object detection allows our method to produce saliency maps that show image areas that most affect the prediction. D-RISE can be considered "black-box" in the software testing sense, as it only needs access to the inputs and outputs of an object detector. Compared to gradient-based methods, D-RISE is more general and agnostic to the particular type of object detector being tested, and does not need knowledge of the inner workings of the model. We show that D-RISE can be easily applied to different object detectors including one-stage detectors such as YOLOv3 and two-stage detectors such as Faster-RCNN. We present a detailed analysis of the generated visual explanations to highlight the utilization of context and possible biases learned by object detectors.

Animals are diverse in shape, but building a deformable shape model for a new species is not always possible due to the lack of 3D data. We present a method to capture new species using an articulated template and images of that species. In this work, we focus mainly on birds. Although birds represent almost twice the number of species as mammals, no accurate shape model is available. To capture a novel species, we first fit the articulated template to each training sample. By disentangling pose and shape, we learn a shape space that captures variation both among species and within each species from image evidence. We learn models of multiple species from the CUB dataset, and contribute new species-specific and multi-species shape models that are useful for downstream reconstruction tasks. Using a low-dimensional embedding, we show that our learned 3D shape space better reflects the phylogenetic relationships among birds than learned perceptual features.

Scene graph generation is an important visual understanding task with a broad range of vision applications. Despite recent tremendous progress, it remains challenging due to the intrinsic long-tailed class distribution and large intra-class variation. To address these issues, we introduce a novel confidence-aware bipartite graph neural network with adaptive message propagation mechanism for unbiased scene graph generation. In addition, we propose an efficient bi-level data resampling strategy to alleviate the imbalanced data distribution problem in training our graph network. Our approach achieves superior or competitive performance over previous methods on several challenging datasets, including Visual Genome, Open Images V4/V6, demonstrating its effectiveness and generality.