ICCV 2019 论文列表

Single-image super-resolution aims to generate a high-resolution version of a low-resolution image, which serves as an essential component in many image processing applications. This paper investigates the robustness of deep learning-based super-resolution methods against adversarial attacks, which can significantly deteriorate the super-resolved images without noticeable distortion in the attacked low-resolution images. It is demonstrated that state-of-the-art deep super-resolution methods are highly vulnerable to adversarial attacks. Different levels of robustness of different methods are analyzed theoretically and experimentally. We also present analysis on transferability of attacks, and feasibility of targeted attacks and universal attacks.

Batch normalization (BN) has been very effective for deep learning and is widely used. However, when training with small minibatches, models using BN exhibit a significant degradation in performance. In this paper we study this peculiar behavior of BN to gain a better understanding of the problem, and identify a cause. We propose `EvalNorm' to address the issue by estimating corrected normalization statistics to use for BN during evaluation. EvalNorm supports online estimation of the corrected statistics while the model is being trained, and does not affect the training scheme of the model. As a result, EvalNorm can also be used with existing pre-trained models allowing them to benefit from our method. EvalNorm yields large gains for models trained with smaller batches. Our experiments show that EvalNorm performs 6.18% (absolute) better than vanilla BN for a batchsize of 2 on ImageNet validation set and from 1.5 to 7.0 points (absolute) gain on the COCO object detection benchmark across a variety of setups.

We present a general method to estimate the fundamental matrix from a pair of images under perspective projection without the need for image point correspondences. Our method is particularly well-suited for transmission imaging, where state-of-the-art feature detection and matching approaches generally do not perform well. Estimation of the fundamental matrix plays a central role in auto-calibration methods for reflection imaging. Such methods are currently not applicable to transmission imaging. Furthermore, our method extends an existing technique proposed for reflection imaging which potentially avoids the outlier-prone feature matching step from an orthographic projection model to a perspective model. Our method exploits the idea that under a linear attenuation model line integrals along corresponding epipolar lines are equal if we compute their derivatives in orthogonal direction to their common epipolar plane. We use the fundamental matrix to parametrize this equality. Our method estimates the matrix by formulating a non-convex optimization problem, minimizing an error in our measurement of this equality. We believe this technique will enable the application of the large body of work on image-based camera pose estimation to transmission imaging leading to more accurate and more general motion compensation and auto-calibration algorithms, particularly in medical X-ray and Computed Tomography imaging.

We introduce PlatonicGAN to discover the 3D structure of an object class from an unstructured collection of 2D images, i.e., where no relation between photos is known, except that they are showing instances of the same category. The key idea is to train a deep neural network to generate 3D shapes which, when rendered to images, are indistinguishable from ground truth images (for a discriminator) under various camera poses. Discriminating 2D images instead of 3D shapes allows tapping into unstructured 2D photo collections instead of relying on curated (e.g., aligned, annotated, etc.) 3D data sets. To establish constraints between 2D image observation and their 3D interpretation, we suggest a family of rendering layers that are effectively differentiable. This family includes visual hull, absorption-only (akin to x-ray), and emission-absorption. We can successfully reconstruct 3D shapes from unstructured 2D images and extensively evaluate PlatonicGAN on a range of synthetic and real data sets achieving consistent improvements over baseline methods. We further show that PlatonicGAN can be combined with 3D supervision to improve on and in some cases even surpass the quality of 3D-supervised methods.

Several popular approaches to 3D vision tasks process multiple views of the input independently with deep neural networks pre-trained on natural images, where view permutation invariance is achieved through a single round of pooling over all views. We argue that this operation discards important information and leads to subpar global descriptors. In this paper, we propose a group convolutional approach to multiple view aggregation where convolutions are performed over a discrete subgroup of the rotation group, enabling, thus, joint reasoning over all views in an equivariant (instead of invariant) fashion, up to the very last layer. We further develop this idea to operate on smaller discrete homogeneous spaces of the rotation group, where a polar view representation is used to maintain equivariance with only a fraction of the number of input views. We set the new state of the art in several large scale 3D shape retrieval tasks, and show additional applications to panoramic scene classification.

Domain generalization (DG) is the challenging and topical problem of learning models that generalize to novel testing domains with different statistics than a set of known training domains. The simple approach of aggregating data from all source domains and training a single deep neural network end-to-end on all the data provides a surprisingly strong baseline that surpasses many prior published methods. In this paper we build on this strong baseline by designing an episodic training procedure that trains a single deep network in a way that exposes it to the domain shift that characterises a novel domain at runtime. Specifically, we decompose a deep network into feature extractor and classifier components, and then train each component by simulating it interacting with a partner who is badly tuned for the current domain. This makes both components more robust, ultimately leading to our networks producing state-of-the-art performance on three DG benchmarks. Furthermore, we consider the pervasive workflow of using an ImageNet trained CNN as a fixed feature extractor for downstream recognition tasks. Using the Visual Decathlon benchmark, we demonstrate that our episodic-DG training improves the performance of such a general purpose feature extractor by explicitly training a feature for robustness to novel problems. This shows that DG training can benefit standard practice in computer vision.

In image captioning, the typical attention mechanisms are arduous to identify the equivalent visual signals especially when predicting highly abstract words. This phenomenon is known as the semantic gap between vision and language. This problem can be overcome by providing semantic attributes that are homologous to language. Thanks to the inherent recurrent nature and gated operating mechanism, Recurrent Neural Network (RNN) and its variants are the dominating architectures in image captioning. However, when designing elaborate attention mechanisms to integrate visual inputs and semantic attributes, RNN-like variants become unflexible due to their complexities. In this paper, we investigate a Transformer-based sequence modeling framework, built only with attention layers and feedforward layers. To bridge the semantic gap, we introduce EnTangled Attention (ETA) that enables the Transformer to exploit semantic and visual information simultaneously. Furthermore, Gated Bilateral Controller (GBC) is proposed to guide the interactions between the multimodal information. We name our model as ETA-Transformer. Remarkably, ETA-Transformer achieves state-of-the-art performance on the MSCOCO image captioning dataset. The ablation studies validate the improvements of our proposed modules.

We propose a single-stage detection framework that jointly tackles the problem of multi-scale object detection and class imbalance. Rather than designing deeper networks, we introduce a simple yet effective feature enrichment scheme to produce multi-scale contextual features. We further introduce a cascaded refinement scheme which first instills multi-scale contextual features into the prediction layers of the single-stage detector in order to enrich their discriminative power for multi-scale detection. Second, the cascaded refinement scheme counters the class imbalance problem by refining the anchors and enriched features to improve classification and regression. Experiments are performed on two benchmarks: PASCAL VOC and MS COCO. For a 320x320 input on the MS COCO test-dev, our detector achieves state-of-the-art single-stage detection accuracy with a COCO AP of 33.2 in the case of single-scale inference, while operating at 21 milliseconds on a Titan XP GPU. For a 512x512 input on the MS COCO test-dev, our approach obtains an absolute gain of 1.6% in terms of COCO AP, compared to the best reported single-stage results[5]. Source code and models are available at: https://github.com/Ranchentx/EFGRNet.

Enhancing low light videos, which consists of denoising and brightness adjustment, is an intriguing but knotty problem. Under low light condition, due to high sensitivity camera setting, commonly negligible noises become obvious and severely deteriorate the captured videos. To recover high quality videos, a mass of image/video denoising/enhancing algorithms are proposed, most of which follow a set of simple assumptions about the statistic characters of camera noise, e.g., independent and identically distributed(i.i.d.), white, additive, Gaussian, Poisson or mixture noises. However, the practical noise under high sensitivity setting in real captured videos is complex and inaccurate to model with these assumptions. In this paper, we explore the physical origins of the practical high sensitivity noise in digital cameras, model them mathematically, and propose to enhance the low light videos based on the noise model by using an LSTM-based neural network. Specifically, we generate the training data with the proposed noise model and train the network with the dark noisy video as input and clear-bright video as output. Extensive comparisons on both synthetic and real captured low light videos with the state-of-the-art methods are conducted to demonstrate the effectiveness of the proposed method.

Neural networks are vulnerable to adversarial examples, malicious inputs crafted to fool trained models. Adversarial examples often exhibit black-box transfer, meaning that adversarial examples for one model can fool another model. However, adversarial examples are typically overfit to exploit the particular architecture and feature representation of a source model, resulting in sub-optimal black-box transfer attacks to other target models. We introduce the Intermediate Level Attack (ILA), which attempts to fine-tune an existing adversarial example for greater black-box transferability by increasing its perturbation on a pre-specified layer of the source model, improving upon state-of-the-art methods. We show that we can select a layer of the source model to perturb without any knowledge of the target models while achieving high transferability. Additionally, we provide some explanatory insights regarding our method and the effect of optimizing for adversarial examples using intermediate feature maps.

Multi-view shape descriptors obtained from various 2D images are commonly adopted in 3D shape retrieval. One major challenge is that significant shape information are discarded during 2D view rendering through projection. In this paper, we propose a convolutional neural network based method, CenterNet, to enhance each individual 2D view using its neighboring ones. By exploiting cross-view correlations, CenterNet learns how adjacent views can be maximally incorporated for an enhanced 2D representation to effectively describe shapes. We observe that a very small amount of, e.g., six, enhanced 2D views, are already sufficient for a panoramic shape description. Thus, by simply aggregating features from six enhanced 2D views, we arrive at a highly compact yet discriminative shape descriptor. The proposed shape descriptor significantly outperforms state-of-the-art 3D shape retrieval methods on the ModelNet and ShapeNetCore55 benchmarks, and also exhibits robustness against object occlusion.

Monocular depth prediction plays a crucial role in understanding 3D scene geometry. Although recent methods have achieved impressive progress in evaluation metrics such as the pixel-wise relative error, most methods neglect the geometric constraints in the 3D space. In this work, we show the importance of the high-order 3D geometric constraints for depth prediction. By designing a loss term that enforces one simple type of geometric constraints, namely, virtual normal directions determined by randomly sampled three points in the reconstructed 3D space, we can considerably improve the depth prediction accuracy. Furthermore, we can not only predict accurate depth but also achieve high-quality other 3D information from the depth without retraining new parameters, Significantly, the byproduct of this predicted depth being sufficiently accurate is that we are now able to recover good 3D structures of the scene such as the point cloud and surface normal directly from the depth, eliminating the necessity of training new sub-models as was previously done. Experiments on two challenging benchmarks: NYU Depth-V2 and KITTI demonstrate the effectiveness of our method and state-of-the-art performance.

We present a conceptually simple yet effective algorithm to detect wireframes in a given image. Compared to the previous methods which first predict an intermediate heat map and then extract straight lines with heuristic algorithms, our method is end-to-end trainable and can directly output a vectorized wireframe that contains semantically meaningful and geometrically salient junctions and lines. To better understand the quality of the outputs, we propose a new metric for wireframe evaluation that penalizes overlapped line segments and incorrect line connectivities. We conduct extensive experiments and show that our method significantly outperforms the previous state-of-the-art wireframe and line extraction algorithms. We hope our simple approach can be served as a baseline for future wireframe parsing studies. Code has been made publicly available at https://github.com/zhou13/lcnn.

Event cameras are vision sensors that record asynchronous streams of per-pixel brightness changes, referred to as "events". They have appealing advantages over frame based cameras for computer vision, including high temporal resolution, high dynamic range, and no motion blur. Due to the sparse, non-uniform spatio-temporal layout of the event signal, pattern recognition algorithms typically aggregate events into a grid-based representation and subsequently process it by a standard vision pipeline, e.g., Convolutional Neural Network (CNN). In this work, we introduce a general framework to convert event streams into grid-based representations by means of strictly differentiable operations. Our framework comes with two main advantages: (i) allows learning the input event representation together with the task dedicated network in an end to end manner, and (ii) lays out a taxonomy that unifies the majority of extant event representations in the literature and identifies novel ones. Empirically, we show that our approach to learning the event representation end-to-end yields an improvement of approximately 12% on optical flow estimation and object recognition over state-of-the-art methods.

Deep neural networks provide powerful tools for pattern recognition, while classical graph algorithms are widely used to solve combinatorial problems. In computer vision, many tasks combine elements of both pattern recognition and graph reasoning. In this paper, we study how to connect deep networks with graph decomposition into an end-to-end trainable framework. More specifically, the minimum cost multicut problem is first converted to an unconstrained binary cubic formulation where cycle consistency constraints are incorporated into the objective function. The new optimization problem can be viewed as a Conditional Random Field (CRF) in which the random variables are associated with the binary edge labels. Cycle constraints are introduced into the CRF as high-order potentials. A standard Convolutional Neural Network (CNN) provides the front-end features for the fully differentiable CRF. The parameters of both parts are optimized in an end-to-end manner. The efficacy of the proposed learning algorithm is demonstrated via experiments on clustering MNIST images and on the challenging task of real-world multi-people pose estimation.

In this paper, we present a HAnd Mesh Recovery (HAMR) framework to tackle the problem of reconstructing the full 3D mesh of a human hand from a single RGB image. In contrast to existing research on 2D or 3D hand pose estimation from RGB or/and depth image data, HAMR can provide a more expressive and useful mesh representation for monocular hand image understanding. In particular, the mesh representation is achieved by parameterizing a generic 3D hand model with shape and relative 3D joint angles. By utilizing this mesh representation, we can easily compute the 3D joint locations via linear interpolations between the vertexes of the mesh, while obtain the 2D joint locations with a projection of the 3D joints. To this end, a differentiable re-projection loss can be defined in terms of the derived representations and the ground-truth labels, thus making our framework end-to-end trainable. Qualitative experiments show that our framework is capable of recovering appealing 3D hand mesh even in the presence of severe occlusions. Quantitatively, our approach also outperforms the state-of-the-art methods for both 2D and 3D hand pose estimation from a monocular RGB image on several benchmark datasets.

We present a novel, end-to-end approach to align CAD models to an 3D scan of a scene, enabling transformation of a noisy, incomplete 3D scan to a compact, CAD reconstruction with clean, complete object geometry. Our main contribution lies in formulating a differentiable Procrustes alignment that is paired with a symmetry-aware dense object correspondence prediction. To simultaneously align CAD models to all the objects of a scanned scene, our approach detects object locations, then predicts symmetry-aware dense object correspondences between scan and CAD geometry in a unified object space, as well as a nearest neighbor CAD model, both of which are then used to inform a differentiable Procrustes alignment. Our approach operates in a fully-convolutional fashion, enabling alignment of CAD models to the objects of a scan in a single forward pass. This enables our method to outperform state-of-the-art approaches by 19.04% for CAD model alignment to scans, with approximately 250x faster runtime than previous data-driven approaches.

Despite Convolutional Neural Networks (CNNs) based methods have been successful in detecting salient objects, their underlying mechanism that decides the salient intensity of each image part separately cannot avoid inconsistency of parts within the same salient object. This would ultimately result in an incomplete shape of the detected salient object. To solve this problem, we dig into part-object relationships and take the unprecedented attempt to employ these relationships endowed by the Capsule Network (CapsNet) for salient object detection. The entire salient object detection system is built directly on a Two-Stream Part-Object Assignment Network (TSPOANet) consisting of three algorithmic steps. In the first step, the learned deep feature maps of the input image are transformed to a group of primary capsules. In the second step, we feed the primary capsules into two identical streams, within each of which low-level capsules (parts) will be assigned to their familiar high-level capsules (object) via a locally connected routing. In the final step, the two streams are integrated in the form of a fully connected layer, where the relevant parts can be clustered together to form a complete salient object. Experimental results demonstrate the superiority of the proposed salient object detection network over the state-of-the-art methods.

Passive visual systems typically fail to recognize objects in the amodal setting where they are heavily occluded. In contrast, humans and other embodied agents have the ability to move in the environment and actively control the viewing angle to better understand object shapes and semantics. In this work, we introduce the task of Embodied Amodel Recognition (EAR): an agent is instantiated in a 3D environment close to an occluded target object, and is free to move in the environment to perform object classification, amodal object localization, and amodal object segmentation. To address this problem, we develop a new model called Embodied Mask R-CNN for agents to learn to move strategically to improve their visual recognition abilities. We conduct experiments using a simulator for indoor environments. Experimental results show that: 1) agents with embodiment (movement) achieve better visual recognition performance than passive ones and 2) in order to improve visual recognition abilities, agents can learn strategic paths that are different from shortest paths.