Depth completion starts from a sparse set of known depth values and estimates the unknown depths for the remaining image pixels. Most methods model this as depth interpolation and erroneously interpolate depth pixels into the empty space between spatially distinct objects, resulting in depth-smearing across occlusion boundaries. Here we propose a multi-hypothesis depth representation that explicitly models both foreground and background depths in the difficult occlusion-boundary regions. Our method can be thought of as performing twin-surface extrapolation, rather than interpolation, in these regions. Next our method fuses these extrapolated surfaces into a single depth image leveraging the image data. Key to our method is the use of an asymmetric loss function that operates on a novel twin-surface representation. This enables us to train a network to simultaneously do surface extrapolation and surface fusion. We characterize our loss function and compare with other common losses. Finally, we validate our method on three different datasets; KITTI, an outdoor real-world dataset, NYU2, indoor real-world depth dataset and Virtual KITTI, a photo-realistic synthetic dataset with dense groundtruth, and demonstrate improvement over the state of the art.
State-of-the-art defense mechanisms against face attacks achieve near perfect accuracies within one of three attack categories, namely adversarial, digital manipulation, or physical spoofs, however, they fail to generalize well when tested across all three categories. Poor generalization can be attributed to learning incoherent attacks jointly. To overcome this shortcoming, we propose a unified attack detection framework, namely UniFAD, that can automatically cluster 25 coherent attack types belonging to the three categories. Using a multi-task learning framework along with k-means clustering, UniFAD learns joint representations for coherent attacks, while uncorrelated attack types are learned separately. Proposed UniFAD outperforms prevailing defense methods and their fusion with an overall TDR = 94.73% @ 0.2% FDR on a large fake face dataset consisting of 341K bona fide images and 448K attack images of 25 types across all 3 categories. Proposed method can detect an attack within 3 milliseconds on a Nvidia 2080Ti. UniFAD can also identify the attack types and categories with 75.81% and 97.37% accuracies, respectively.
Inferring 3D structure of a generic object from a 2D image is a long-standing objective of computer vision. Conventional approaches either learn completely from CAD-generated synthetic data, which have difficulty in inference from real images, or generate 2.5D depth image via intrinsic decomposition, which is limited compared to the full 3D reconstruction. One fundamental challenge lies in how to leverage numerous real 2D images without any 3D ground truth. To address this issue, we take an alternative approach with semi-supervised learning. That is, for a 2D image of a generic object, we decompose it into latent representations of category, shape and albedo, lighting and camera projection matrix, decode the representations to segmented 3D shape and albedo respectively, and fuse these components to render an image well approximating the input image. Using a category-adaptive 3D joint occupancy field (JOF), we show that the complete shape and albedo modeling enables us to leverage real 2D images in both modeling and model fitting. The effectiveness of our approach is demonstrated through superior 3D reconstruction from a single image, being either synthetic or real, and shape segmentation.
Existing face relighting methods often struggle with two problems: maintaining the local facial details of the subject and accurately removing and synthesizing shadows in the relit image, especially hard shadows. We propose a novel deep face relighting method that addresses both problems. Our method learns to predict the ratio (quotient) image between a source image and the target image with the desired lighting, allowing us to relight the image while maintaining the local facial details. During training, our model also learns to accurately modify shadows by using estimated shadow masks to emphasize on the high-contrast shadow borders. Furthermore, we introduce a method to use the shadow mask to estimate the ambient light intensity in an image, and are thus able to leverage multiple datasets during training with different global lighting intensities. With quantitative and qualitative evaluations on the Multi-PIE and FFHQ datasets, we demonstrate that our proposed method faithfully maintains the local facial details of the subject and can accurately handle hard shadows while achieving state-of-the-art face relighting performance.
Modern 3D object detectors have immensely benefited from the end-to-end learning idea. However, most of them use a post-processing algorithm called Non-Maximal Suppression (NMS) only during inference. While there were attempts to include NMS in the training pipeline for tasks such as 2D object detection, they have been less widely adopted due to a non-mathematical expression of the NMS. In this paper, we present and integrate GrooMeD-NMS -- a novel Grouped Mathematically Differentiable NMS for monocular 3D object detection, such that the network is trained end-to-end with a loss on the boxes after NMS. We first formulate NMS as a matrix operation and then group and mask the boxes in an unsupervised manner to obtain a simple closed-form expression of the NMS. GrooMeD-NMS addresses the mismatch between training and inference pipelines and, therefore, forces the network to select the best 3D box in a differentiable manner. As a result, GrooMeD-NMS achieves state-of-the-art monocular 3D object detection results on the KITTI benchmark dataset performing comparably to monocular video-based methods. Code and models at https://github.com/abhi1kumar/groomed_nms
To defend against manipulation of image content, such as splicing, copy-move, and removal, we develop a Progressive Spatio-Channel Correlation Network (PSCC-Net) to detect and localize image manipulations. PSCC-Net processes the image in a two-path procedure: a top-down path that extracts local and global features and a bottom-up path that detects whether the input image is manipulated, and estimates its manipulation masks at 4 scales, where each mask is conditioned on the previous one. Different from the conventional encoder-decoder and no-pooling structures, PSCC-Net leverages features at different scales with dense cross-connections to produce manipulation masks in a coarse-to-fine fashion. Moreover, a Spatio-Channel Correlation Module (SCCM) captures both spatial and channel-wise correlations in the bottom-up path, which endows features with holistic cues, enabling the network to cope with a wide range of manipulation attacks. Thanks to the light-weight backbone and progressive mechanism, PSCC-Net can process 1,080P images at 50+ FPS. Extensive experiments demonstrate the superiority of PSCC-Net over the state-of-the-art methods on both detection and localization.
Facing the sparsity of user attributes on social networks, attribute inference aims at inferring missing attributes based on existing data and additional information such as social connections between users. Recently, Variational Autoencoders (VAEs) have been successfully applied to solve the problem in a semi-supervised way. However, the latent representations learned by the encoder contain either insufficient or useless information: i) MLPs can successfully reconstruct the input data but fail in completing missing part, ii) GNNs merge information according to social connections but suffer from over-smoothing, which is a common problem with GNNs. Moreover, existing methods neglect regulating the decoder, as a result, it lacks adequate inference ability and faces severe overfitting. To address the above issues, we propose an attribute inference model based on adversarial VAE (Infer-AVAE). Our model deliberately unifies MLPs and GNNs in encoder to learn dual latent representations: one contains only the observed attributes of each user, the other converges extra information from the neighborhood. Then, an adversarial network is trained to leverage the differences between the two representations and adversarial training is conducted to guide GNNs using MLPs for robust representations. What's more, mutual information constraint is introduced in loss function to specifically train the decoder as a discriminator. Thus, it can make better use of auxiliary information in the representations for attribute inference. Based on real-world social network datasets, experimental results demonstrate that our model averagely outperforms state-of-art by 7.0% in accuracy.
Existing surface registration methods focus on fitting in-sample data with little to no generalization ability and require both heavy pre-processing and careful hand-tuning. In this paper, we cast the registration task as a surface-to-surface translation problem, and design a model to reliably capture the latent geometric information directly from raw 3D face scans. We introduce Shape-My-Face (SMF), a powerful encoder-decoder architecture based on an improved point cloud encoder, a novel visual attention mechanism, graph convolutional decoders with skip connections, and a specialized mouth model that we smoothly integrate with the mesh convolutions. Compared to the previous state-of-the-art learning algorithms for non-rigid registration of face scans, SMF only requires the raw data to be rigidly aligned (with scaling) with a pre-defined face template. Additionally, our model provides topologically-sound meshes with minimal supervision, offers faster training time, has orders of magnitude fewer trainable parameters, is more robust to noise, and can generalize to previously unseen datasets. We extensively evaluate the quality of our registrations on diverse data. We demonstrate the robustness and generalizability of our model with in-the-wild face scans across different modalities, sensor types, and resolutions. Finally, we show that, by learning to register scans, SMF produces a hybrid linear and non-linear morphable model that can be used for generation, shape morphing, and expression transfer through manipulation of the latent space, including in-the-wild. We train SMF on a dataset of human faces comprising 9 large-scale databases on commodity hardware.
Prior studies show that the key to face anti-spoofing lies in the subtle image pattern, termed "spoof trace", e.g., color distortion, 3D mask edge, Moire pattern, and many others. Designing a generic face anti-spoofing model to estimate those spoof traces can improve not only the generalization of the spoof detection, but also the interpretability of the model's decision. Yet, this is a challenging task due to the diversity of spoof types and the lack of ground truth in spoof traces. In this work, we design a novel adversarial learning framework to disentangle spoof faces into the spoof traces and the live counterparts. Guided by physical properties, the spoof generation is represented as a combination of additive process and inpainting process. Additive process describes spoofing as spoof material introducing extra patterns (e.g., moire pattern), where the live counterpart can be recovered by removing those patterns. Inpainting process describes spoofing as spoof material fully covering certain regions, where the live counterpart of those regions has to be "guessed". We use 3 additive components and 1 inpainting component to represent traces at different frequency bands. The disentangled spoof traces can be utilized to synthesize realistic new spoof faces after proper geometric correction, and the synthesized spoof can be used for training and improve the generalization of spoof detection. Our approach demonstrates superior spoof detection performance on 3 testing scenarios: known attacks, unknown attacks, and open-set attacks. Meanwhile, it provides a visually-convincing estimation of the spoof traces. Source code and pre-trained models will be publicly available upon publication.
Knowledge graph (KG), as the side information, is widely utilized to learn the semantic representations of item/user for recommendation system. The traditional recommendation algorithms usually just depend on user-item interactions, but ignore the inherent web information describing the item/user, which could be formulated by the knowledge graph embedding (KGE) methods to significantly improve applications' performance. In this paper, we propose a knowledge-aware-based recommendation algorithm to capture the local and global representation learning from heterogeneous information. Specifically, the local model and global model can naturally depict the inner patterns in the content-based heterogeneous information and interactive behaviors among the users and items. Based on the method that local and global representations are learned jointly by graph convolutional networks with attention mechanism, the final recommendation probability is calculated by a fully-connected neural network. Extensive experiments are conducted on two real-world datasets to verify the proposed algorithm's validation. The evaluation results indicate that the proposed algorithm surpasses state-of-arts by $10.0\%$, $5.1\%$, $2.5\%$ and $1.8\%$ in metrics of MAE, RMSE, AUC and F1-score at least, respectively. The significant improvements reveal the capacity of our proposal to recommend user/item effectively.