Hui Zhang

Peng Xiang, Xin Wen, Yu-Shen Liu et al. · tsinghua

Learning per-point semantic features from the hierarchical feature pyramid is essential for point cloud semantic segmentation. However, most previous methods suffered from ambiguous region features or failed to refine per-point features effectively, which leads to information loss and ambiguous semantic identification. To resolve this, we propose Retro-FPN to model the per-point feature prediction as an explicit and retrospective refining process, which goes through all the pyramid layers to extract semantic features explicitly for each point. Its key novelty is a retro-transformer for summarizing semantic contexts from the previous layer and accordingly refining the features in the current stage. In this way, the categorization of each point is conditioned on its local semantic pattern. Specifically, the retro-transformer consists of a local cross-attention block and a semantic gate unit. The cross-attention serves to summarize the semantic pattern retrospectively from the previous layer. And the gate unit carefully incorporates the summarized contexts and refines the current semantic features. Retro-FPN is a pluggable neural network that applies to hierarchical decoders. By integrating Retro-FPN with three representative backbones, including both point-based and voxel-based methods, we show that Retro-FPN can significantly improve performance over state-of-the-art backbones. Comprehensive experiments on widely used benchmarks can justify the effectiveness of our design. The source is available at https://github.com/AllenXiangX/Retro-FPN

Hui Zhang, Zheng Wang, Dan Zeng et al.

Anomaly detection has garnered extensive applications in real industrial manufacturing due to its remarkable effectiveness and efficiency. However, previous generative-based models have been limited by suboptimal reconstruction quality, hampering their overall performance. We introduce DiffusionAD, a novel anomaly detection pipeline comprising a reconstruction sub-network and a segmentation sub-network. A fundamental enhancement lies in our reformulation of the reconstruction process using a diffusion model into a noise-to-norm paradigm. Here, the anomalous region loses its distinctive features after being disturbed by Gaussian noise and is subsequently reconstructed into an anomaly-free one. Afterward, the segmentation sub-network predicts pixel-level anomaly scores based on the similarities and discrepancies between the input image and its anomaly-free reconstruction. Additionally, given the substantial decrease in inference speed due to the iterative denoising nature of diffusion models, we revisit the denoising process and introduce a rapid one-step denoising paradigm. This paradigm achieves hundreds of times acceleration while preserving comparable reconstruction quality. Furthermore, considering the diversity in the manifestation of anomalies, we propose a norm-guided paradigm to integrate the benefits of multiple noise scales, enhancing the fidelity of reconstructions. Comprehensive evaluations on four standard and challenging benchmarks reveal that DiffusionAD outperforms current state-of-the-art approaches and achieves comparable inference speed, demonstrating the effectiveness and broad applicability of the proposed pipeline. Code is released at https://github.com/HuiZhang0812/DiffusionAD

Yihong Cao, Hui Zhang, Xiao Lu et al.

Domain adaptive semantic segmentation enables robust pixel-wise understanding in real-world driving scenes. Source-free domain adaptation, as a more practical technique, addresses the concerns of data privacy and storage limitations in typical unsupervised domain adaptation methods, making it especially relevant in the context of intelligent vehicles. It utilizes a well-trained source model and unlabeled target data to achieve adaptation in the target domain. However, in the absence of source data and target labels, current solutions cannot sufficiently reduce the impact of domain shift and fully leverage the information from the target data. In this paper, we propose an end-to-end source-free domain adaptation semantic segmentation method via Importance-Aware and Prototype-Contrast (IAPC) learning. The proposed IAPC framework effectively extracts domain-invariant knowledge from the well-trained source model and learns domain-specific knowledge from the unlabeled target domain. Specifically, considering the problem of domain shift in the prediction of the target domain by the source model, we put forward an importance-aware mechanism for the biased target prediction probability distribution to extract domain-invariant knowledge from the source model. We further introduce a prototype-contrast strategy, which includes a prototype-symmetric cross-entropy loss and a prototype-enhanced cross-entropy loss, to learn target intra-domain knowledge without relying on labels. A comprehensive variety of experiments on two domain adaptive semantic segmentation benchmarks demonstrates that the proposed end-to-end IAPC solution outperforms existing state-of-the-art methods. The source code is publicly available at https://github.com/yihong-97/Source-free-IAPC.

Hui Zhang, Quanming Yao, James T. Kwok et al. · tsinghua

Feature extractor plays a critical role in text recognition (TR), but customizing its architecture is relatively less explored due to expensive manual tweaking. In this work, inspired by the success of neural architecture search (NAS), we propose to search for suitable feature extractors. We design a domain-specific search space by exploring principles for having good feature extractors. The space includes a 3D-structured space for the spatial model and a transformed-based space for the sequential model. As the space is huge and complexly structured, no existing NAS algorithms can be applied. We propose a two-stage algorithm to effectively search in the space. In the first stage, we cut the space into several blocks and progressively train each block with the help of an auxiliary head. We introduce the latency constraint into the second stage and search sub-network from the trained supernet via natural gradient descent. In experiments, a series of ablation studies are performed to better understand the designed space, search algorithm, and searched architectures. We also compare the proposed method with various state-of-the-art ones on both hand-written and scene TR tasks. Extensive results show that our approach can achieve better recognition performance with less latency.

Hui Zhang, Wotao Yin

The convergence behavior of gradient methods for minimizing convex differentiable functions is one of the core questions in convex optimization. This paper shows that their well-known complexities can be achieved under conditions weaker than the commonly accepted ones. We relax the common gradient Lipschitz-continuity condition and strong convexity condition to ones that hold only over certain line segments. Specifically, we establish complexities $O(\frac{R}ε)$ and $O(\sqrt{\frac{R}ε})$ for the ordinary and accelerate gradient methods, respectively, assuming that $\nabla f$ is Lipschitz continuous with constant $R$ over the line segment joining $x$ and $x-\frac{1}{R}\nabla f$ for each $x\in\dom f$. Then we improve them to $O(\frac{R}ν\log(\frac{1}ε))$ and $O(\sqrt{\frac{R}ν}\log(\frac{1}ε))$ for function $f$ that also satisfies the secant inequality $\ < \nabla f(x), x- x^*\ > \ge ν\|x-x^*\|^2$ for each $x\in \dom f$ and its projection $x^*$ to the minimizer set of $f$. The secant condition is also shown to be necessary for the geometric decay of solution error. Not only are the relaxed conditions met by more functions, the restrictions give smaller $R$ and larger $ν$ than they are without the restrictions and thus lead to better complexity bounds. We apply these results to sparse optimization and demonstrate a faster algorithm.

Hui Zhang, Zuxuan Wu, Zheng Wang et al.

Anomaly detection and localization are widely used in industrial manufacturing for its efficiency and effectiveness. Anomalies are rare and hard to collect and supervised models easily over-fit to these seen anomalies with a handful of abnormal samples, producing unsatisfactory performance. On the other hand, anomalies are typically subtle, hard to discern, and of various appearance, making it difficult to detect anomalies and let alone locate anomalous regions. To address these issues, we propose a framework called Prototypical Residual Network (PRN), which learns feature residuals of varying scales and sizes between anomalous and normal patterns to accurately reconstruct the segmentation maps of anomalous regions. PRN mainly consists of two parts: multi-scale prototypes that explicitly represent the residual features of anomalies to normal patterns; a multisize self-attention mechanism that enables variable-sized anomalous feature learning. Besides, we present a variety of anomaly generation strategies that consider both seen and unseen appearance variance to enlarge and diversify anomalies. Extensive experiments on the challenging and widely used MVTec AD benchmark show that PRN outperforms current state-of-the-art unsupervised and supervised methods. We further report SOTA results on three additional datasets to demonstrate the effectiveness and generalizability of PRN.

Hui Zhang, Shenglong Zhou, Geoffrey Ye Li et al.

The step function is one of the simplest and most natural activation functions for deep neural networks (DNNs). As it counts 1 for positive variables and 0 for others, its intrinsic characteristics (e.g., discontinuity and no viable information of subgradients) impede its development for several decades. Even if there is an impressive body of work on designing DNNs with continuous activation functions that can be deemed as surrogates of the step function, it is still in the possession of some advantageous properties, such as complete robustness to outliers and being capable of attaining the best learning-theoretic guarantee of predictive accuracy. Hence, in this paper, we aim to train DNNs with the step function used as an activation function (dubbed as 0/1 DNNs). We first reformulate 0/1 DNNs as an unconstrained optimization problem and then solve it by a block coordinate descend (BCD) method. Moreover, we acquire closed-form solutions for sub-problems of BCD as well as its convergence properties. Furthermore, we also integrate $\ell_{2,0}$-regularization into 0/1 DNN to accelerate the training process and compress the network scale. As a result, the proposed algorithm has a high performance on classifying MNIST and Fashion-MNIST datasets. As a result, the proposed algorithm has a desirable performance on classifying MNIST, FashionMNIST, Cifar10, and Cifar100 datasets.

Zhen Xing, Qi Dai, Zihao Zhang et al.

Diffusion models have achieved significant success in image and video generation. This motivates a growing interest in video editing tasks, where videos are edited according to provided text descriptions. However, most existing approaches only focus on video editing for short clips and rely on time-consuming tuning or inference. We are the first to propose Video Instruction Diffusion (VIDiff), a unified foundation model designed for a wide range of video tasks. These tasks encompass both understanding tasks (such as language-guided video object segmentation) and generative tasks (video editing and enhancement). Our model can edit and translate the desired results within seconds based on user instructions. Moreover, we design an iterative auto-regressive method to ensure consistency in editing and enhancing long videos. We provide convincing generative results for diverse input videos and written instructions, both qualitatively and quantitatively. More examples can be found at our website https://ChenHsing.github.io/VIDiff.

Hui Zhang, Zuxuan Wu, Zhen Xing et al.

Diffusion models, as a type of generative model, have achieved impressive results in generating images and videos conditioned on textual conditions. However, the generation process of diffusion models involves denoising dozens of steps to produce photorealistic images/videos, which is computationally expensive. Unlike previous methods that design ``one-size-fits-all'' approaches for speed up, we argue denoising steps should be sample-specific conditioned on the richness of input texts. To this end, we introduce AdaDiff, a lightweight framework designed to learn instance-specific step usage policies, which are then used by the diffusion model for generation. AdaDiff is optimized using a policy gradient method to maximize a carefully designed reward function, balancing inference time and generation quality. We conduct experiments on three image generation and two video generation benchmarks and demonstrate that our approach achieves similar visual quality compared to the baseline using a fixed 50 denoising steps while reducing inference time by at least 33%, going as high as 40%. Furthermore, our method can be used on top of other acceleration methods to provide further speed benefits. Lastly, qualitative analysis shows that AdaDiff allocates more steps to more informative prompts and fewer steps to simpler prompts.

Zhanhao Hu, Wenda Chu, Xiaopei Zhu et al.

Recent works have proposed to craft adversarial clothes for evading person detectors, while they are either only effective at limited viewing angles or very conspicuous to humans. We aim to craft adversarial texture for clothes based on 3D modeling, an idea that has been used to craft rigid adversarial objects such as a 3D-printed turtle. Unlike rigid objects, humans and clothes are non-rigid, leading to difficulties in physical realization. In order to craft natural-looking adversarial clothes that can evade person detectors at multiple viewing angles, we propose adversarial camouflage textures (AdvCaT) that resemble one kind of the typical textures of daily clothes, camouflage textures. We leverage the Voronoi diagram and Gumbel-softmax trick to parameterize the camouflage textures and optimize the parameters via 3D modeling. Moreover, we propose an efficient augmentation pipeline on 3D meshes combining topologically plausible projection (TopoProj) and Thin Plate Spline (TPS) to narrow the gap between digital and real-world objects. We printed the developed 3D texture pieces on fabric materials and tailored them into T-shirts and trousers. Experiments show high attack success rates of these clothes against multiple detectors.

Xiaoshuai Hao, Yunfeng Diao, Mengchuan Wei et al.

Map construction task plays a vital role in providing precise and comprehensive static environmental information essential for autonomous driving systems. Primary sensors include cameras and LiDAR, with configurations varying between camera-only, LiDAR-only, or camera-LiDAR fusion, based on cost-performance considerations. While fusion-based methods typically perform best, existing approaches often neglect modality interaction and rely on simple fusion strategies, which suffer from the problems of misalignment and information loss. To address these issues, we propose MapFusion, a novel multi-modal Bird's-Eye View (BEV) feature fusion method for map construction. Specifically, to solve the semantic misalignment problem between camera and LiDAR BEV features, we introduce the Cross-modal Interaction Transform (CIT) module, enabling interaction between two BEV feature spaces and enhancing feature representation through a self-attention mechanism. Additionally, we propose an effective Dual Dynamic Fusion (DDF) module to adaptively select valuable information from different modalities, which can take full advantage of the inherent information between different modalities. Moreover, MapFusion is designed to be simple and plug-and-play, easily integrated into existing pipelines. We evaluate MapFusion on two map construction tasks, including High-definition (HD) map and BEV map segmentation, to show its versatility and effectiveness. Compared with the state-of-the-art methods, MapFusion achieves 3.6% and 6.2% absolute improvements on the HD map construction and BEV map segmentation tasks on the nuScenes dataset, respectively, demonstrating the superiority of our approach.

Junbang Liu, Enpei Huang, Dongxing Mao et al.

Creating 3D content from single-view images is a challenging problem that has attracted considerable attention in recent years. Current approaches typically utilize score distillation sampling (SDS) from pre-trained 2D diffusion models to generate multi-view 3D representations. Although some methods have made notable progress by balancing generation speed and model quality, their performance is often limited by the visual inconsistencies of the diffusion model outputs. In this work, we propose ContrastiveGaussian, which integrates contrastive learning into the generative process. By using a perceptual loss, we effectively differentiate between positive and negative samples, leveraging the visual inconsistencies to improve 3D generation quality. To further enhance sample differentiation and improve contrastive learning, we incorporate a super-resolution model and introduce another Quantity-Aware Triplet Loss to address varying sample distributions during training. Our experiments demonstrate that our approach achieves superior texture fidelity and improved geometric consistency.

Xiaoshuai Hao, Mengchuan Wei, Yifan Yang et al.

Driving systems often rely on high-definition (HD) maps for precise environmental information, which is crucial for planning and navigation. While current HD map constructors perform well under ideal conditions, their resilience to real-world challenges, \eg, adverse weather and sensor failures, is not well understood, raising safety concerns. This work introduces MapBench, the first comprehensive benchmark designed to evaluate the robustness of HD map construction methods against various sensor corruptions. Our benchmark encompasses a total of 29 types of corruptions that occur from cameras and LiDAR sensors. Extensive evaluations across 31 HD map constructors reveal significant performance degradation of existing methods under adverse weather conditions and sensor failures, underscoring critical safety concerns. We identify effective strategies for enhancing robustness, including innovative approaches that leverage multi-modal fusion, advanced data augmentation, and architectural techniques. These insights provide a pathway for developing more reliable HD map construction methods, which are essential for the advancement of autonomous driving technology. The benchmark toolkit and affiliated code and model checkpoints have been made publicly accessible.

Ping Zhang, Zhenxiang Tao, Wenjie Yang et al.

Surveillance cameras are widely applied for indoor occupancy measurement and human movement perception, which benefit for building energy management and social security. To address the challenges of limited view angle of single camera as well as lacking of inter-camera collaboration, this study presents a non-overlapping multi-camera system to enlarge the surveillance area and devotes to retrieve the same person appeared from different camera views. The system is deployed in an office building and four-day videos are collected. By training a deep convolutional neural network, the proposed system first extracts the appearance feature embeddings of each personal image, which detected from different cameras, for similarity comparison. Then, a stochastic inter-camera transition matrix is associated with appearance feature for further improving the person re-identification ranking results. Finally, a noise-suppression explanation is given for analyzing the matching improvements. This paper expands the scope of indoor movement perception based on non-overlapping multiple cameras and improves the accuracy of pedestrian re-identification without introducing additional types of sensors.

Wandong Zhang, QM Jonathan Wu, Yimin Yang et al.

Most multilayer least squares (LS)-based neural networks are structured with two separate stages: unsupervised feature encoding and supervised pattern classification. Once the unsupervised learning is finished, the latent encoding would be fixed without supervised fine-tuning. However, in complex tasks such as handling the ImageNet dataset, there are often many more clues that can be directly encoded, while the unsupervised learning, by definition cannot know exactly what is useful for a certain task. This serves as the motivation to retrain the latent space representations to learn some clues that unsupervised learning has not yet learned. In particular, the error matrix from the output layer is pulled back to each hidden layer, and the parameters of the hidden layer are recalculated with Moore-Penrose (MP) inverse for more generalized representations. In this paper, a recomputation-based multilayer network using MP inverse (RML-MP) is developed. A sparse RML-MP (SRML-MP) model to boost the performance of RML-MP is then proposed. The experimental results with varying training samples (from 3 K to 1.8 M) show that the proposed models provide better generalization performance than most representation learning algorithms.

Xinzhi Wang, Huao Li, Hui Zhang et al.

Recently, there has been increasing interest in transparency and interpretability in Deep Reinforcement Learning (DRL) systems. Verbal explanations, as the most natural way of communication in our daily life, deserve more attention, since they allow users to gain a better understanding of the system which ultimately could lead to a high level of trust and smooth collaboration. This paper reports a novel work in generating verbal explanations for DRL behaviors agent. A rule-based model is designed to construct explanations using a series of rules which are predefined with prior knowledge. A learning model is then proposed to expand the implicit logic of generating verbal explanation to general situations by employing rule-based explanations as training data. The learning model is shown to have better flexibility and generalizability than the static rule-based model. The performance of both models is evaluated quantitatively through objective metrics. The results show that verbal explanation generated by both models improve subjective satisfaction of users towards the interpretability of DRL systems. Additionally, seven variants of the learning model are designed to illustrate the contribution of input channels, attention mechanism, and proposed encoder in improving the quality of verbal explanation.

Hui Zhang, Quanming Yao, Mingkun Yang et al.

Scene text recognition (STR) is very challenging due to the diversity of text instances and the complexity of scenes. The community has paid increasing attention to boost the performance by improving the pre-processing image module, like rectification and deblurring, or the sequence translator. However, another critical module, i.e., the feature sequence extractor, has not been extensively explored. In this work, inspired by the success of neural architecture search (NAS), which can identify better architectures than human-designed ones, we propose automated STR (AutoSTR) to search data-dependent backbones to boost text recognition performance. First, we design a domain-specific search space for STR, which contains both choices on operations and constraints on the downsampling path. Then, we propose a two-step search algorithm, which decouples operations and downsampling path, for an efficient search in the given space. Experiments demonstrate that, by searching data-dependent backbones, AutoSTR can outperform the state-of-the-art approaches on standard benchmarks with much fewer FLOPS and model parameters.

Peng Liu, Hui Zhang, Kie B. Eom

Active deep learning classification of hyperspectral images is considered in this paper. Deep learning has achieved success in many applications, but good-quality labeled samples are needed to construct a deep learning network. It is expensive getting good labeled samples in hyperspectral images for remote sensing applications. An active learning algorithm based on a weighted incremental dictionary learning is proposed for such applications. The proposed algorithm selects training samples that maximize two selection criteria, namely representative and uncertainty. This algorithm trains a deep network efficiently by actively selecting training samples at each iteration. The proposed algorithm is applied for the classification of hyperspectral images, and compared with other classification algorithms employing active learning. It is shown that the proposed algorithm is efficient and effective in classifying hyperspectral images.

Zhuangming Shen, Jiuai Sun, Hui Zhang et al.

Joint saliency map (JSM) [1] was developed to assign high joint saliency values to the corresponding saliency structures (called Joint Saliency Structures, JSSs) but zero or low joint saliency values to the outliers (or mismatches) that are introduced by missing correspondence or local large deformations between the reference and moving images to be registered. JSM guides the local structure matching in nonrigid registration by emphasizing these JSSs' sparse deformation vectors in adaptive kernel regression of hierarchical sparse deformation vectors for iterative dense deformation reconstruction. By designing an effective superpixel-based local structure scale estimator to compute the reference structure's structure scale, we further propose to determine the scale (the width) of kernels in the adaptive kernel regression through combining the structure scales to JSM-based scales of mismatch between the local saliency structures. Therefore, we can adaptively select the sample size of sparse deformation vectors to reconstruct the dense deformation vectors for accurately matching the every local structures in the two images. The experimental results demonstrate better accuracy of our method in aligning two images with missing correspondence and local large deformation than the state-of-the-art methods.

Binjie Qin, Zhuangming Shen, Zien Zhou et al.

For nonrigid image registration, matching the particular structures (or the outliers) that have missing correspondence and/or local large deformations, can be more difficult than matching the common structures with small deformations in the two images. Most existing works depend heavily on the outlier segmentation to remove the outlier effect in the registration. Moreover, these works do not handle simultaneously the missing correspondences and local large deformations. In this paper, we defined the nonrigid image registration as a local adaptive kernel regression which locally reconstruct the moving image's dense deformation vectors from the sparse deformation vectors in the multi-resolution block matching. The kernel function of the kernel regression adapts its shape and orientation to the reference image's structure to gather more deformation vector samples of the same structure for the iterative regression computation, whereby the moving image's local deformations could be compliant with the reference image's local structures. To estimate the local deformations around the outliers, we use joint saliency map that highlights the corresponding saliency structures (called Joint Saliency Structures, JSSs) in the two images to guide the dense deformation reconstruction by emphasizing those JSSs' sparse deformation vectors in the kernel regression. The experimental results demonstrate that by using local JSS adaptive kernel regression, the proposed method achieves almost the best performance in alignment of all challenging image pairs with outlier structures compared with other five state-of-the-art nonrigid registration algorithms.