CVPR 2019 Papers

Letting a deep network be aware of the quality of its own predictions is an interesting yet important problem. In the task of instance segmentation, the confidence of instance classification is used as mask quality score in most instance segmentation frameworks. However, the mask quality, quantified as the IoU between the instance mask and its ground truth, is usually not well correlated with classification score. In this paper, we study this problem and propose Mask Scoring R-CNN which contains a network block to learn the quality of the predicted instance masks. The proposed network block takes the instance feature and the corresponding predicted mask together to regress the mask IoU. The mask scoring strategy calibrates the misalignment between mask quality and mask score, and improves instance segmentation performance by prioritizing more accurate mask predictions during COCO AP evaluation. By extensive evaluations on the COCO dataset, Mask Scoring R-CNN brings consistent and noticeable gain with different models and outperforms the state-of-the-art Mask R-CNN. We hope our simple and effective approach will provide a new direction for improving instance segmentation. The source code of our method is available at https://github.com/zjhuang22/maskscoring_rcnn.

Zero-shot learning has been well explored to precisely identify new unobserved classes through a visual-semantic function obtained from the existing objects. However, there exist two challenging obstacles: one is that the human-annotated semantics are insufficient to fully describe the visual samples; the other is the domain shift across existing and new classes. In this paper, we attempt to exploit the intrinsic relationship in the semantic manifold when given semantics are not enough to describe the visual objects, and enhance the generalization ability of the visual-semantic function with marginalized strategy. Specifically, we design a Marginalized Latent Semantic Encoder (MLSE), which is learned on the augmented seen visual features and the latent semantic representation. Meanwhile, latent semantics are discovered under an adaptive graph reconstruction scheme based on the provided semantics. Consequently, our proposed algorithm could enrich visual characteristics from seen classes, and well generalize to unobserved classes. Experimental results on zero-shot benchmarks demonstrate that the proposed model delivers superior performance over the state-of-the-art zero-shot learning approaches.

Building on progress in feature representations for image retrieval, image-based localization has seen a surge of research interest. Image-based localization has the advantage of being inexpensive and efficient, often avoiding the use of 3D metric maps altogether. That said, the need to maintain a large amount of reference images as an effective support of localization in a scene, nonetheless calls for them to be organized in a map structure of some kind. The problem of localization often arises as part of a navigation process. We are, therefore, interested in summarizing the reference images as a set of landmarks, which meet the requirements for image-based navigation. A contribution of this paper is to formulate such a set of requirements for the two sub-tasks involved: compact map construction and accurate self localization. These requirements are then exploited for compact map representation and accurate self-localization, using the framework of a network flow problem. During this process, we formulate the map construction and self-localization problems as convex quadratic and second-order cone programs, respectively. We evaluate our methods on publicly available indoor and outdoor datasets, where they outperform existing methods significantly.

To fight against real-life image forgery, which commonly involves different types and combined manipulations, we propose a unified deep neural architecture called ManTra-Net. Unlike many existing solutions, ManTra-Net is an end-to-end network that performs both detection and localization without extra preprocessing and postprocessing. \manifold is a fully convolutional network and handles images of arbitrary sizes and many known forgery types such splicing, copy-move, removal, enhancement, and even unknown types. This paper has three salient contributions. We design a simple yet effective self-supervised learning task to learn robust image manipulation traces from classifying 385 image manipulation types. Further, we formulate the forgery localization problem as a local anomaly detection problem, design a Z-score feature to capture local anomaly, and propose a novel long short-term memory solution to assess local anomalies. Finally, we carefully conduct ablation experiments to systematically optimize the proposed network design. Our extensive experimental results demonstrate the generalizability, robustness and superiority of ManTra-Net, not only in single types of manipulations/forgeries, but also in their complicated combinations.

Cloud based medical image analysis has become popular recently due to the high computation complexities of various deep neural network (DNN) based frameworks and the increasingly large volume of medical images that need to be processed. It has been demonstrated that for medical images the transmission from local to clouds is much more expensive than the computation in the clouds itself. Towards this, 3D image compression techniques have been widely applied to reduce the data traffic. However, most of the existing image compression techniques are developed around human vision, i.e., they are designed to minimize distortions that can be perceived by human eyes. In this paper, we will use deep learning based medical image segmentation as a vehicle and demonstrate that interestingly, machine and human view the compression quality differently. Medical images compressed with good quality w.r.t. human vision may result in inferior segmentation accuracy. We then design a machine vision oriented 3D image compression framework tailored for segmentation using DNNs. Our method automatically extracts and retains image features that are most important to the segmentation. Comprehensive experiments on widely adopted segmentation frameworks with HVSMR 2016 challenge dataset show that our method can achieve significantly higher segmentation accuracy at the same compression rate, or much better compression rate under the same segmentation accuracy, when compared with the existing JPEG 2000 method. To the best of the authors' knowledge, this is the first machine vision guided medical image compression framework for segmentation in the clouds.

The detection of anomalous structures in natural image data is of utmost importance for numerous tasks in the field of computer vision. The development of methods for unsupervised anomaly detection requires data on which to train and evaluate new approaches and ideas. We introduce the MVTec Anomaly Detection (MVTec AD) dataset containing 5354 high-resolution color images of different object and texture categories. It contains normal, i.e., defect-free, images intended for training and images with anomalies intended for testing. The anomalies manifest themselves in the form of over 70 different types of defects such as scratches, dents, contaminations, and various structural changes. In addition, we provide pixel-precise ground truth regions for all anomalies. We also conduct a thorough evaluation of current state-of-the-art unsupervised anomaly detection methods based on deep architectures such as convolutional autoencoders, generative adversarial networks, and feature descriptors using pre-trained convolutional neural networks, as well as classical computer vision methods. This initial benchmark indicates that there is considerable room for improvement. To the best of our knowledge, this is the first comprehensive, multi-object, multi-defect dataset for anomaly detection that provides pixel-accurate ground truth regions and focuses on real-world applications.

We address the problem of recovering the 3D geometry of a human face from a set of facial images in multiple views. While recent studies have shown impressive progress in 3D Morphable Model (3DMM) based facial reconstruction, the settings are mostly restricted to a single view. There is an inherent drawback in the single-view setting: the lack of reliable 3D constraints can cause unresolvable ambiguities. We in this paper explore 3DMM-based shape recovery in a different setting, where a set of multi-view facial images are given as input. A novel approach is proposed to regress 3DMM parameters from multi-view inputs with an end-to-end trainable Convolutional Neural Network (CNN). Multi-view geometric constraints are incorporated into the network by establishing dense correspondences between different views leveraging a novel self-supervised view alignment loss. The main ingredient of the view alignment loss is a differentiable dense optical flow estimator that can backpropagate the alignment errors between an input view and a synthetic rendering from another input view, which is projected to the target view through the 3D shape to be inferred. Through minimizing the view alignment loss, better 3D shapes can be recovered such that the synthetic projections from one view to another can better align with the observed image. Extensive experiments demonstrate the superiority of the proposed method over other 3DMM methods.

Multimodal attentional networks are currently state-of-the-art models for Visual Question Answering (VQA) tasks involving real images. Although attention allows to focus on the visual content relevant to the question, this simple mechanism is arguably insufficient to model complex reasoning features required for VQA or other high-level tasks. In this paper, we propose MuRel, a multimodal relational network which is learned end-to-end to reason over real images. Our first contribution is the introduction of the MuRel cell, an atomic reasoning primitive representing interactions between question and image regions by a rich vectorial representation, and modeling region relations with pairwise combinations. Secondly, we incorporate the cell into a full MuRel network, which progressively refines visual and question interactions, and can be leveraged to define visualization schemes finer than mere attention maps. We validate the relevance of our approach with various ablation studies, and show its superiority to attention-based methods on three datasets: VQA 2.0, VQA-CP v2 and TDIUC. Our final MuRel network is competitive to or outperforms state-of-the-art results in this challenging context. Our code is available: github.com/Cadene/murel.bootstrap.pytorch

In this paper, we propose an adversarial learning network for the task of multi-style image captioning (MSCap) with a standard factual image caption dataset and a multi-stylized language corpus without paired images. How to learn a single model for multi-stylized image captioning with unpaired data is a challenging and necessary task, whereas rarely studied in previous works. The proposed framework mainly includes four contributive modules following a typical image encoder. First, a style dependent caption generator to output a sentence conditioned on an encoded image and a specified style. Second, a caption discriminator is presented to distinguish the input sentence to be real or not. The discriminator and the generator are trained in an adversarial manner to enable more natural and human-like captions. Third, a style classifier is employed to discriminate the specific style of the input sentence. Besides, a back-translation module is designed to enforce the generated stylized captions are visually grounded, with the intuition of the cycle consistency for factual caption and stylized caption. We enable an end-to-end optimization of the whole model with differentiable softmax approximation. At last, we conduct comprehensive experiments using a combined dataset containing four caption styles to demonstrate the outstanding performance of our proposed method.

Temporally locating and classifying action segments in long untrimmed videos is of particular interest to many applications like surveillance and robotics. While traditional approaches follow a two-step pipeline, by generating frame-wise probabilities and then feeding them to high-level temporal models, recent approaches use temporal convolutions to directly classify the video frames. In this paper, we introduce a multi-stage architecture for the temporal action segmentation task. Each stage features a set of dilated temporal convolutions to generate an initial prediction that is refined by the next one. This architecture is trained using a combination of a classification loss and a proposed smoothing loss that penalizes over-segmentation errors. Extensive evaluation shows the effectiveness of the proposed model in capturing long-range dependencies and recognizing action segments. Our model achieves state-of-the-art results on three challenging datasets: 50Salads, Georgia Tech Egocentric Activities (GTEA), and the Breakfast dataset.

This paper extends the popular task of multi-object tracking to multi-object tracking and segmentation (MOTS). Towards this goal, we create dense pixel-level annotations for two existing tracking datasets using a semi-automatic annotation procedure. Our new annotations comprise 65,213 pixel masks for 977 distinct objects (cars and pedestrians) in 10,870 video frames. For evaluation, we extend existing multi-object tracking metrics to this new task. Moreover, we propose a new baseline method which jointly addresses detection, tracking, and segmentation with a single convolutional network. We demonstrate the value of our datasets by achieving improvements in performance when training on MOTS annotations. We believe that our datasets, metrics and baseline will become a valuable resource towards developing multi-object tracking approaches that go beyond 2D bounding boxes. We make our annotations, code, and models available at https://www.vision.rwth-aachen.de/page/mots.

We propose to address the face reconstruction in the wild by using a multi-metric regression network, MMFace, to align a 3D face morphable model (3DMM) to an input image. The key idea is to utilize a volumetric sub-network to estimate an intermediate geometry representation, and a parametric sub-network to regress the 3DMM parameters. Our parametric sub-network consists of identity loss, expression loss, and pose loss which greatly improves the aligned geometry details by incorporating high level loss functions directly defined in the 3DMM parametric spaces. Our high-quality reconstruction is robust under large variations of expressions, poses, illumination conditions, and even with large partial occlusions. We evaluate our method by comparing the performance with state-of-the-art approaches on latest 3D face dataset LS3D-W and Florence. We achieve significant improvements both quantitatively and qualitatively. Due to our high-quality reconstruction, our method can be easily extended to generate high-quality geometry sequences for video inputs.

We address the problem of semi-supervised video object segmentation (VOS), where the masks of objects of interests are given in the first frame of an input video. To deal with challenging cases where objects are occluded or missing, previous work relies on greedy data association strategies that make decisions for each frame individually. In this paper, we propose a novel approach to defer the decision making for a target object in each frame, until a global view can be established with the entire video being taken into consideration. Our approach is in the same spirit as Multiple Hypotheses Tracking (MHT) methods, making several critical adaptations for the VOS problem. We employ the bounding box (bbox) hypothesis for tracking tree formation, and the multiple hypotheses are spawned by propagating the preceding bbox into the detected bbox proposals within a gated region starting from the initial object mask in the first frame. The gated region is determined by a gating scheme which takes into account a more comprehensive motion model rather than the simple Kalman filtering model in traditional MHT. To further design more customized algorithms tailored for VOS, we develop a novel mask propagation score instead of the appearance similarity score that could be brittle due to large deformations. The mask propagation score, together with the motion score, determines the affinity between the hypotheses during tree pruning. Finally, a novel mask merging strategy is employed to handle mask conflicts between objects. Extensive experiments on challenging datasets demonstrate the effectiveness of the proposed method, especially in the case of object missing.

We tackle the problem of finding good architectures for multimodal classification problems. We propose a novel and generic search space that spans a large number of possible fusion architectures. In order to find an optimal architecture for a given dataset in the proposed search space, we leverage an efficient sequential model-based exploration approach that is tailored for the problem. We demonstrate the value of posing multimodal fusion as a neural architecture search problem by extensive experimentation on a toy dataset and two other real multimodal datasets. We discover fusion architectures that exhibit state-of-the-art performance for problems with different domain and dataset size, including the \ntu dataset, the largest multimodal action recognition dataset available.

In this paper we propose the macroblock scaling (MBS) algorithm, which can be applied to various CNN architectures to reduce their model size. MBS adaptively reduces each CNN macroblock depending on its information redundancy measured by our proposed effective flops. Empirical studies conducted with ImageNet and CIFAR-10 attest that MBS can reduce the model size of some already compact CNN models, e.g., MobileNetV2 (25.03% further reduction) and ShuffleNet (20.74%), and even ultra-deep ones such as ResNet-101 (51.67%) and ResNet-1202 (72.71%) with negligible accuracy degradation. MBS also performs better reduction at a much lower cost than the state-of-the-art optimization-based methods do. MBS's simplicity and efficiency, its flexibility to work with any CNN model, and its scalability to work with models of any depth make it an attractive choice for CNN model size reduction.

Most state-of-the-art methods for action recognition consist of a two-stream architecture with 3D convolutions: an appearance stream for RGB frames and a motion stream for optical flow frames. Although combining flow with RGB improves the performance, the cost of computing accurate optical flow is high, and increases action recognition latency. This limits the usage of two-stream approaches in real-world applications requiring low latency. In this paper, we introduce two learning approaches to train a standard 3D CNN, operating on RGB frames, that mimics the motion stream, and as a result avoids flow computation at test time. First, by minimizing a feature-based loss compared to the Flow stream, we show that the network reproduces the motion stream with high fidelity. Second, to leverage both appearance and motion information effectively, we train with a linear combination of the feature-based loss and the standard cross-entropy loss for action recognition. We denote the stream trained using this combined loss as Motion-Augmented RGB Stream (MARS). As a single stream, MARS performs better than RGB or Flow alone, for instance with 72.7% accuracy on Kinetics compared to 72.0% and 65.6% with RGB and Flow streams respectively.

We present a new proximal bundle method for Maximum-A-Posteriori (MAP) inference in structured energy minimization problems. The method optimizes a Lagrangean relaxation of the original energy minimization problem using a multi plane block-coordinate Frank-Wolfe method that takes advantage of the specific structure of the Lagrangean decomposition. We show empirically that our method outperforms state-of-the-art Lagrangean decomposition based algorithms on some challenging Markov Random Field, multi-label discrete tomography and graph matching problems.

This research strives for natural language moment retrieval in long, untrimmed video streams. The problem is not trivial especially when a video contains multiple moments of interests and the language describes complex temporal dependencies, which often happens in real scenarios. We identify two crucial challenges: semantic misalignment and structural misalignment. However, existing approaches treat different moments separately and do not explicitly model complex moment-wise temporal relations. In this paper, we present Moment Alignment Network (MAN), a novel framework that unifies the candidate moment encoding and temporal structural reasoning in a single-shot feed-forward network. MAN naturally assigns candidate moment representations aligned with language semantics over different temporal locations and scales. Most importantly, we propose to explicitly model moment-wise temporal relations as a structured graph and devise an iterative graph adjustment network to jointly learn the best structure in an end-to-end manner. We evaluate the proposed approach on two challenging public benchmarks DiDeMo and Charades-STA, where our MAN significantly outperforms the state-of-the-art by a large margin.

A method called, sigma-consensus, is proposed to eliminate the need for a user-defined inlier-outlier threshold in RANSAC. Instead of estimating the noise sigma, it is marginalized over a range of noise scales. The optimized model is obtained by weighted least-squares fitting where the weights come from the marginalization over sigma of the point likelihoods of being inliers. A new quality function is proposed not requiring sigma and, thus, a set of inliers to determine the model quality. Also, a new termination criterion for RANSAC is built on the proposed marginalization approach. Applying sigma-consensus, MAGSAC is proposed with no need for a user-defined sigma and improving the accuracy of robust estimation significantly. It is superior to the state-of-the-art in terms of geometric accuracy on publicly available real-world datasets for epipolar geometry (F and E) and homography estimation. In addition, applying sigma-consensus only once as a post-processing step to the RANSAC output always improved the model quality on a wide range of vision problems without noticeable deterioration in processing time, adding a few milliseconds.