Bo Zhang

Bin Wang, Chao Xu, Xiaomeng Zhao et al.

Document content analysis has been a crucial research area in computer vision. Despite significant advancements in methods such as OCR, layout detection, and formula recognition, existing open-source solutions struggle to consistently deliver high-quality content extraction due to the diversity in document types and content. To address these challenges, we present MinerU, an open-source solution for high-precision document content extraction. MinerU leverages the sophisticated PDF-Extract-Kit models to extract content from diverse documents effectively and employs finely-tuned preprocessing and postprocessing rules to ensure the accuracy of the final results. Experimental results demonstrate that MinerU consistently achieves high performance across various document types, significantly enhancing the quality and consistency of content extraction. The MinerU open-source project is available at https://github.com/opendatalab/MinerU.

Xingxing Wei, Ying Guo, Jie Yu et al.

Adversarial patch is an important form of real-world adversarial attack that brings serious risks to the robustness of deep neural networks. Previous methods generate adversarial patches by either optimizing their perturbation values while fixing the pasting position or manipulating the position while fixing the patch's content. This reveals that the positions and perturbations are both important to the adversarial attack. For that, in this paper, we propose a novel method to simultaneously optimize the position and perturbation for an adversarial patch, and thus obtain a high attack success rate in the black-box setting. Technically, we regard the patch's position, the pre-designed hyper-parameters to determine the patch's perturbations as the variables, and utilize the reinforcement learning framework to simultaneously solve for the optimal solution based on the rewards obtained from the target model with a small number of queries. Extensive experiments are conducted on the Face Recognition (FR) task, and results on four representative FR models show that our method can significantly improve the attack success rate and query efficiency. Besides, experiments on the commercial FR service and physical environments confirm its practical application value. We also extend our method to the traffic sign recognition task to verify its generalization ability.

Canyu Zhao, Mingyu Liu, Wen Wang et al.

Recent advancements in video generation have primarily leveraged diffusion models for short-duration content. However, these approaches often fall short in modeling complex narratives and maintaining character consistency over extended periods, which is essential for long-form video production like movies. We propose MovieDreamer, a novel hierarchical framework that integrates the strengths of autoregressive models with diffusion-based rendering to pioneer long-duration video generation with intricate plot progressions and high visual fidelity. Our approach utilizes autoregressive models for global narrative coherence, predicting sequences of visual tokens that are subsequently transformed into high-quality video frames through diffusion rendering. This method is akin to traditional movie production processes, where complex stories are factorized down into manageable scene capturing. Further, we employ a multimodal script that enriches scene descriptions with detailed character information and visual style, enhancing continuity and character identity across scenes. We present extensive experiments across various movie genres, demonstrating that our approach not only achieves superior visual and narrative quality but also effectively extends the duration of generated content significantly beyond current capabilities. Homepage: https://aim-uofa.github.io/MovieDreamer/.

Bo Zhang, Yuchen Guo, Runzhao Yang et al.

Imaging and perception in photon-limited scenarios is necessary for various applications, e.g., night surveillance or photography, high-speed photography, and autonomous driving. In these cases, cameras suffer from low signal-to-noise ratio, which degrades the image quality severely and poses challenges for downstream high-level vision tasks like object detection and recognition. Data-driven methods have achieved enormous success in both image restoration and high-level vision tasks. However, the lack of high-quality benchmark dataset with task-specific accurate annotations for photon-limited images/videos delays the research progress heavily. In this paper, we contribute the first multi-illuminance, multi-camera, and low-light dataset, named DarkVision, serving for both image enhancement and object detection. We provide bright and dark pairs with pixel-wise registration, in which the bright counterpart provides reliable reference for restoration and annotation. The dataset consists of bright-dark pairs of 900 static scenes with objects from 15 categories, and 32 dynamic scenes with 4-category objects. For each scene, images/videos were captured at 5 illuminance levels using three cameras of different grades, and average photons can be reliably estimated from the calibration data for quantitative studies. The static-scene images and dynamic videos respectively contain around 7,344 and 320,667 instances in total. With DarkVision, we established baselines for image/video enhancement and object detection by representative algorithms. To demonstrate an exemplary application of DarkVision, we propose two simple yet effective approaches for improving performance in video enhancement and object detection respectively. We believe DarkVision would advance the state-of-the-arts in both imaging and related computer vision tasks in low-light environment.

Qingyuan Li, Bo Zhang, Xiangxiang Chu

Structured pruning greatly eases the deployment of large neural networks in resource-constrained environments. However, current methods either involve strong domain expertise, require extra hyperparameter tuning, or are restricted only to a specific type of network, which prevents pervasive industrial applications. In this paper, we undertake a simple and effective approach that can be easily applied to both vision transformers and convolutional neural networks. Specifically, we consider pruning as an evolution process of sub-network structures that inherit weights through reconstruction techniques. We achieve a 50% FLOPS reduction for ResNet50 and MobileNetV1, leading to 1.37x and 1.34x speedup respectively. For DeiT-Base, we reach nearly 40% FLOPs reduction and 1.4x speedup. Our code will be made available.

Haiwen Zhang, Bo Zhang

This paper is concerned with the direct and inverse acoustic or electromagnetic scattering problems by a locally perturbed, perfectly reflecting, infinite plane (which is called a locally rough surface in this paper). We propose a novel integral equation formulation for the direct scattering problem which is defined on a bounded curve (consisting of a bounded part of the infinite plane containing the local perturbation and the lower part of a circle) with two corners. This novel integral equation can be solved efficiently by using the Nystrom method with a graded mesh introduced previously by Kress and is capable of dealing with large wavenumber cases. For the inverse problem, we propose a Newton iteration method to reconstruct the local perturbation of the plane from multiple frequency far-field data, based on the novel integral equation formulation. Numerical examples are carried out to demonstrate that our reconstruction method is stable and accurate even for the case of multiple-scale profiles.

Xiaoli Liu, Bo Zhang, Haiwen Zhang

This paper is concerned with the inverse scattering problem by an unbounded rough surface. A direct imaging method is proposed to reconstruct the rough surface from the scattered near-field Cauchy data generating by point sources and measured on a horizontal straight line segment at a finite distance above the rough surface. Theoretical analysis of the imaging algorithm is given for the case of a penetrable rough surface, but the imaging algorithm also works for impenetrable surfaces with Dirichlet or impedance boundary conditions. Numerical experiments are presented to show that the direct imaging algorithm is fast, accurate and very robust with respect to noise in the data.

Qiao Hu, Bo Zhang, Haiwen Zhang

The contrast source inversion (CSI) method and the subspace-based optimization method (SOM) are first proposed in 1997 and 2009, respectively, and subsequently modified. The two methods and their variants share several properties and thus are called the CSI-type methods. The CSI-type methods are efficient and popular methods for solving inverse medium scattering problems, but their rigorous convergence remains an open problem. In this paper, we propose two iteratively regularized CSI-type (IRCSI-type) methods with a novel $\ell_1$ proximal term as the iteratively regularized term: the iteratively regularized CSI (IRCSI) method and the iteratively regularized SOM (IRSOM) method, which have a similar computation complexity to the original CSI and SOM methods, respectively, and prove their global convergence under natural and weak conditions on the original objective function. To the best of our knowledge, this is the first convergence result for iterative methods of solving nonlinear inverse scattering problems with a fixed frequency. The convergence and performance of the two IRCSI-type algorithms are illustrated by numerical experiments.

Xiaoli Liu, Bo Zhang, Haiwen Zhang

This paper is concerned with the inverse scattering problem of time-harmonic elastic waves by an unbounded rigid rough surface. A direct imaging method is developed to reconstruct the unbounded rough surface from the elastic scattered near-field Cauchy data generated by point sources. A Helmholtz-Kirchhoff-type identity is derived and then used to provide a theoretical analysis of the direct imaging algorithm. Numerical experiments are presented to show that the direct imaging algorithm is fast, accurate and robust with respect to noise in the data.

Guanghui Hu, Xiaoli Liu, Bo Zhang et al.

This paper is concerned with the inverse time-harmonic elastic scattering problem of recovering unbounded rough surfaces in two dimensions. We assume that elastic plane waves with different directions are incident onto a rigid rough surface in a half plane. The elastic scattered field is measured on a horizontal straight line segment within a finite distance above the rough surface. A direct imaging algorithm is proposed to recover the unbounded rough surface from the scattered near-field data, which involves only inner products between the data. Numerical experiments are presented to show that the inversion scheme is not only efficient but also accurate and robust with respect to noise.

Lihong Wang, Liangqi Li, Weiwei Feng et al.

CoT has significantly enhanced the reasoning ability of LLMs while it faces challenges when extended to multimodal domains, particularly in mathematical tasks. Existing MLLMs typically perform textual reasoning solely from a single static mathematical image, overlooking dynamic visual acquisition during reasoning. In contrast, humans repeatedly examine visual image and employ step-by-step reasoning to prove intermediate propositions. This strategy of decomposing the problem-solving process into key logical nodes adheres to Miller's Law in cognitive science. Inspired by this insight, we propose a ViRC framework for multimodal mathematical tasks, introducing a Reason Chunking mechanism that structures multimodal mathematical CoT into consecutive Critical Reasoning Units (CRUs) to simulate human expert problem-solving patterns. CRUs ensure intra-unit textual coherence for intermediate proposition verification while integrating visual information across units to generate subsequent propositions and support structured reasoning. To this end, we present CRUX dataset by using three visual tools and four reasoning patterns to provide explicitly annotated CRUs across multiple reasoning paths for each mathematical problem. Leveraging the CRUX dataset, we propose a progressive training strategy inspired by human cognitive learning, which includes Instructional SFT, Practice SFT, and Strategic RL, aimed at further strengthening the Reason Chunking ability of the model. The resulting ViRC-7B model achieves a 18.8% average improvement over baselines across multiple mathematical benchmarks. Code is available at https://github.com/Leon-LihongWang/ViRC.

Difei Cheng, Yunfeng Zhang, Ruinan Jin

K-medoids clustering is a popular variant of k-means clustering and widely used in pattern recognition and machine learning. A main drawback of k-medoids clustering is that an improper initialization can cause it to get trapped in local optima. An improved k-medoids clustering algorithm, called INCKM algorithm, which is the first to apply incremental initialization to k-medoids clustering, was recently proposed to overcome this drawback. The INCKM algorithm requires the construction of a subset of candidate medoids determined by one hyperparameter for initialization, and meanwhile, it always fails when dealing with imbalanced datasets with an incorrect hyperparameter selection. In this paper, we propose a novel k-medoids clustering algorithm, called incremental k-means++ (INCKPP) algorithm, which initializes with a novel incremental manner, attempting to optimally add one new cluster center at each stage through a nonparametric and stochastic k-means++ initialization. The INCKPP algorithm overcomes the difficulty of hyperparameter selection in the INCKM algorithm, improves the clustering performance, and can deal with imbalanced datasets well. However, the INCKPP algorithm is not computationally efficient enough. To deal with this, we further propose an improved INCKPP algorithm, called INCKPPsample algorithm, which improves the clustering efficiency while maintaining the clustering performance of the INCKPP algorithm. Extensive results from experiments on both synthetic and real-world datasets, including imbalanced datasets, illustrate that the proposed algorithms outperforms than the other compared algorithms.

Difei Cheng, Ruihang Xu, Bo Zhang et al.

Density-based clustering algorithms are widely used for discovering clusters in pattern recognition and machine learning since they can deal with non-hyperspherical clusters and are robustness to handle outliers. However, the runtime of density-based algorithms are heavily dominated by finding fixed-radius near neighbors and calculating the density, which is time-consuming. Meanwhile, the traditional acceleration methods using indexing technique such as KD tree is not effective in processing high-dimensional data. In this paper, we propose a fast region query algorithm named fast principal component analysis pruning (called FPCAP) with the help of the fast principal component analysis technique in conjunction with geometric information provided by principal attributes of the data, which can process high-dimensional data and be easily applied to density-based methods to prune unnecessary distance calculations when finding neighbors and estimating densities. As an application in density-based clustering methods, FPCAP method was combined with the Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. And then, an improved DBSCAN (called IDBSCAN) is obtained, which preserves the advantage of DBSCAN and meanwhile, greatly reduces the computation of redundant distances. Experiments on seven benchmark datasets demonstrate that the proposed algorithm improves the computational efficiency significantly.

Fan Bao, Guoqiang Wu, Chongxuan Li et al.

The (gradient-based) bilevel programming framework is widely used in hyperparameter optimization and has achieved excellent performance empirically. Previous theoretical work mainly focuses on its optimization properties, while leaving the analysis on generalization largely open. This paper attempts to address the issue by presenting an expectation bound w.r.t. the validation set based on uniform stability. Our results can explain some mysterious behaviours of the bilevel programming in practice, for instance, overfitting to the validation set. We also present an expectation bound for the classical cross-validation algorithm. Our results suggest that gradient-based algorithms can be better than cross-validation under certain conditions in a theoretical perspective. Furthermore, we prove that regularization terms in both the outer and inner levels can relieve the overfitting problem in gradient-based algorithms. In experiments on feature learning and data reweighting for noisy labels, we corroborate our theoretical findings.

Junru Luo, Hong Qiao, Bo Zhang

Imbalanced classification tasks are widespread in many real-world applications. For such classification tasks, in comparison with the accuracy rate, it is usually much more appropriate to use non-decomposable performance measures such as the Area Under the receiver operating characteristic Curve (AUC) and the $F_β$ measure as the classification criterion since the label class is imbalanced. On the other hand, the minimax probability machine is a popular method for binary classification problems and aims at learning a linear classifier by maximizing the accuracy rate, which makes it unsuitable to deal with imbalanced classification tasks. The purpose of this paper is to develop a new minimax probability machine for the $F_β$ measure, called MPMF, which can be used to deal with imbalanced classification tasks. A brief discussion is also given on how to extend the MPMF model for several other non-decomposable performance measures listed in the paper. To solve the MPMF model effectively, we derive its equivalent form which can then be solved by an alternating descent method to learn a linear classifier. Further, the kernel trick is employed to derive a nonlinear MPMF model to learn a nonlinear classifier. Several experiments on real-world benchmark datasets demonstrate the effectiveness of our new model.

JunRu Luo, Hong Qiao, Bo Zhang

Due to the non-smoothness of the Hinge loss in SVM, it is difficult to obtain a faster convergence rate with modern optimization algorithms. In this paper, we introduce two smooth Hinge losses $ψ_G(α;σ)$ and $ψ_M(α;σ)$ which are infinitely differentiable and converge to the Hinge loss uniformly in $α$ as $σ$ tends to $0$. By replacing the Hinge loss with these two smooth Hinge losses, we obtain two smooth support vector machines(SSVMs), respectively. Solving the SSVMs with the Trust Region Newton method (TRON) leads to two quadratically convergent algorithms. Experiments in text classification tasks show that the proposed SSVMs are effective in real-world applications. We also introduce a general smooth convex loss function to unify several commonly-used convex loss functions in machine learning. The general framework provides smooth approximation functions to non-smooth convex loss functions, which can be used to obtain smooth models that can be solved with faster convergent optimization algorithms.

Minfang Lu, Shuai Ning, Shuangrong Liu et al.

Black-box optimization (BBO) algorithms are concerned with finding the best solutions for problems with missing analytical details. Most classical methods for such problems are based on strong and fixed a priori assumptions, such as Gaussianity. However, the complex real-world problems, especially when the global optimum is desired, could be very far from the a priori assumptions because of their diversities, causing unexpected obstacles. In this study, we propose a generative adversarial net-based broad-spectrum global optimizer (OPT-GAN) which estimates the distribution of optimum gradually, with strategies to balance exploration-exploitation trade-off. It has potential to better adapt to the regularity and structure of diversified landscapes than other methods with fixed prior, e.g., Gaussian assumption or separability. Experiments on diverse BBO benchmarks and high dimensional real world applications exhibit that OPT-GAN outperforms other traditional and neural net-based BBO algorithms.

Fengfu Li, Hong Qiao, Bo Zhang et al.

Traditional image clustering methods take a two-step approach, feature learning and clustering, sequentially. However, recent research results demonstrated that combining the separated phases in a unified framework and training them jointly can achieve a better performance. In this paper, we first introduce fully convolutional auto-encoders for image feature learning and then propose a unified clustering framework to learn image representations and cluster centers jointly based on a fully convolutional auto-encoder and soft $k$-means scores. At initial stages of the learning procedure, the representations extracted from the auto-encoder may not be very discriminative for latter clustering. We address this issue by adopting a boosted discriminative distribution, where high score assignments are highlighted and low score ones are de-emphasized. With the gradually boosted discrimination, clustering assignment scores are discriminated and cluster purities are enlarged. Experiments on several vision benchmark datasets show that our methods can achieve a state-of-the-art performance.

Fengfu Li, Xiayuan Huang, Hong Qiao et al.

Manifold distances are very effective tools for visual object recognition. However, most of the traditional manifold distances between images are based on the pixel-level comparison and thus easily affected by image rotations and translations. In this paper, we propose a new manifold distance to model the dissimilarities between visual objects based on the Complex Wavelet Structural Similarity (CW-SSIM) index. The proposed distance is more robust to rotations and translations of images than the traditional manifold distance and the CW-SSIM index based distance. In addition, the proposed distance is combined with the $k$-medoids clustering method to derive a new clustering method for visual object categorization. Experiments on Coil-20, Coil-100 and Olivetti Face Databases show that the proposed distance measure is better for visual object categorization than both the traditional manifold distances and the CW-SSIM index based distances.

Hugh Perkins, Minjie Xu, Jun Zhu et al.

As one of the most popular classifiers, linear SVMs still have challenges in dealing with very large-scale problems, even though linear or sub-linear algorithms have been developed recently on single machines. Parallel computing methods have been developed for learning large-scale SVMs. However, existing methods rely on solving local sub-optimization problems. In this paper, we develop a novel parallel algorithm for learning large-scale linear SVM. Our approach is based on a data augmentation equivalent formulation, which casts the problem of learning SVM as a Bayesian inference problem, for which we can develop very efficient parallel sampling methods. We provide empirical results for this parallel sampling SVM, and provide extensions for SVR, non-linear kernels, and provide a parallel implementation of the Crammer and Singer model. This approach is very promising in its own right, and further is a very useful technique to parallelize a broader family of general maximum-margin models.

Wenbo Hu, Jun Zhu, Bo Zhang

Bayesian max-margin models have shown superiority in various practical applications, such as text categorization, collaborative prediction, social network link prediction and crowdsourcing, and they conjoin the flexibility of Bayesian modeling and predictive strengths of max-margin learning. However, Monte Carlo sampling for these models still remains challenging, especially for applications that involve large-scale datasets. In this paper, we present the stochastic subgradient Hamiltonian Monte Carlo (HMC) methods, which are easy to implement and computationally efficient. We show the approximate detailed balance property of subgradient HMC which reveals a natural and validated generalization of the ordinary HMC. Furthermore, we investigate the variants that use stochastic subsampling and thermostats for better scalability and mixing. Using stochastic subgradient Markov Chain Monte Carlo (MCMC), we efficiently solve the posterior inference task of various Bayesian max-margin models and extensive experimental results demonstrate the effectiveness of our approach.

Ning Chen, Jun Zhu, Jianfei Chen et al.

Dropout and other feature noising schemes have shown promising results in controlling over-fitting by artificially corrupting the training data. Though extensive theoretical and empirical studies have been performed for generalized linear models, little work has been done for support vector machines (SVMs), one of the most successful approaches for supervised learning. This paper presents dropout training for linear SVMs. To deal with the intractable expectation of the non-smooth hinge loss under corrupting distributions, we develop an iteratively re-weighted least square (IRLS) algorithm by exploring data augmentation techniques. Our algorithm iteratively minimizes the expectation of a re-weighted least square problem, where the re-weights have closed-form solutions. The similar ideas are applied to develop a new IRLS algorithm for the expected logistic loss under corrupting distributions. Our algorithms offer insights on the connection and difference between the hinge loss and logistic loss in dropout training. Empirical results on several real datasets demonstrate the effectiveness of dropout training on significantly boosting the classification accuracy of linear SVMs.

Xiaodong Liu, Bo Zhang

This paper is concerned with the inverse problem of scattering of time-harmonic acoustic waves by an inhomogeneous penetrable obstacle in a piecewise homogeneous medium. The well-posedness of the direct problem is first established by using the integral equation method. We then proceed to establish two tools that play an important role for the inverse problem: one is a mixed reciprocity relation and the other is a priori estimates of the solution on some part of the interfaces between the layered media. For the inverse problem, we prove in this paper that both the penetrable interfaces and the possible inside inhomogeneity can be uniquely determined from a knowledge of the far field pattern for incident plane waves.