ZZZeyu Zhang

Zeyu Zhang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Zeyu Zhang has a total of 27 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    X-VoE: Measuring eXplanatory Violation of Expectation in Physical Events

    by Bo Dai, Linge Wang, Baoxiong Jia, Zeyu Zhang, Song-Chun Zhu, Chi Zhang, Yixin Zhu

    Intuitive physics is pivotal for human understanding of the physical world, enabling prediction and interpretation of events even in infancy. Nonetheless, replicating this level of intuitive physics in artificial intelligence (AI) remains a formidable challenge. This study introduces X-VoE, a comprehensive benchmark dataset, to assess AI agents’ grasp of intuitive physics. Built on the developmental psychology-rooted Violation of Expectation (VoE) paradigm, X-VoE establishes a higher bar for the explanatory capacities of intuitive physics models. Each VoE scenario within X-VoE encompasses three distinct settings, probing models’ comprehension of events and their underlying explanations. Beyond model evaluation, we present an explanation-based learning system that captures physics dynamics and infers occluded object states solely from visual sequences, without explicit occlusion labels. Experimental outcomes highlight our model’s alignment with human commonsense when tested against X-VoE. A remarkable feature is our model’s ability to visually expound VoE events by reconstructing concealed scenes. Concluding, we discuss the findings’ implications and outline future research directions. Through X-VoE, we catalyze the advancement of AI endowed with human-like intuitive physics capabilities.

  • Open Access

    An Anti-Noise Fast Circle Detection Method Using Five-Quadrant Segmentation

    by Yun Ou, Honggui Deng, Yang Liu, Zeyu Zhang, Xin Lan

    Extracting circle information from images has always been a basic problem in computer vision. Common circle detection algorithms have some defects, such as poor noise resistance and slow computation speed. In this paper, we propose an anti-noise fast circle detection algorithm. In order to improve the anti-noise of the algorithm, we first perform curve thinning and connection on the image after edge extraction, then suppress noise interference by the irregularity of noise edges and extract circular arcs by directional filtering. In order to reduce the invalid fitting and speed up the running speed, we propose a circle fitting algorithm with five quadrants, and improve the efficiency of the algorithm by the idea of “divide and conquer”. We compare the algorithm with RCD, CACD, WANG and AS on two open datasets. The results show that we have the best performance under noise while keeping the speed of the algorithm.


  • Open Access

    WDRASS: A Web-scale Dataset for Document Retrieval and Answer Sentence Selection

    by Zeyu Zhang, Thuy Vu, Sunil Gandhi, Ankit Chadha, Alessandro Moschitti

    Open-Domain Question Answering (ODQA) systems generate answers from relevant text returned by search engines, e.g., lexical features-based such as BM25, or embeddings-based such as dense passage retrieval (DPR). Few datasets are available for this task: they mainly focus on QA systems based on machine reading (MR) approach, and show problematic evaluation, mostly based on uncontextualized short answer matching. In this paper, we present WDRASS, a dataset for ODQA based on answer sentence selection (AS2) models, which consider sentences as candidate answers for QA systems. WDRASS consists of ∼64k questions and 800k+ labeled passages and sentences extracted from 30M documents. We evaluate the dataset by training models on it and comparing with the same models trained on Google NQ. Our experiments show that WDRASS significantly improves the performance of retrieval and reranking models, thus boosting the accuracy of downstream QA tasks. We believe our dataset can produce significant impact in advancing IR research.


  • Open Access

    A Fast Circle Detection Algorithm Based on Information Compression

    by Yun Ou, Honggui Deng, Yang Liu, Zeyu Zhang, Xusheng Ruan, Qiguo Xu, Chengzuo Peng

    Circle detection is a fundamental problem in computer vision. However, conventional circle detection algorithms are usually time-consuming and sensitive to noise. In order to solve these shortcomings, we propose a fast circle detection algorithm based on information compression. First, we introduce the idea of information compression, which compresses the circular information on the image into a small number of points while removing some of the noise through sharpness estimation and orientation filtering. Then, the circle parameters stored in the information point are obtained by the average sampling algorithm with a time complexity of O(1) to obtain candidate circles. Finally, we set different constraints on the complete circle and the defective circle according to the sampling results and find the true circle from the candidate circles. The experimental results on the three datasets show that our method can compress the circular information in the image into 1% of the information points, and compared to RHT, RCD, Jiang, Wang and CACD, Precision, Recall, Time and F-measure are greatly improved.


  • Open Access

    A ConvNext-Based and Feature Enhancement Anchor-Free Siamese Network for Visual Tracking

    by Qiguo Xu, Honggui Deng, Zeyu Zhang, Yang Liu, Xusheng Ruan, Gang Liu

    Existing anchor-based Siamese trackers rely on the anchor’s design to predict the scale and aspect ratio of the target. However, these methods introduce many hyperparameters, leading to computational redundancy. In this paper, to achieve outstanding network efficiency, we propose a ConvNext-based anchor-free Siamese tracking network (CAFSN), which employs an anchor-free design to increase network flexibility and versatility. In CAFSN, to obtain an appropriate backbone network, the state-of-the-art ConvNext network is applied to the visual tracking field for the first time by improving the network stride and receptive field. Moreover, A central confidence branch based on Euclidean distance is offered to suppress low-quality prediction frames in the classification prediction network of CAFSN for robust visual tracking. In particular, we discuss that the Siamese network cannot establish a complete identification model for the tracking target and similar objects, which negatively impacts network performance. We build a Fusion network consisting of crop and 3Dmaxpooling to better distinguish the targets and similar objects’ abilities. This module uses 3DMaxpooling to select the highest activation value to improve the difference between it and other similar objects. Crop unifies the dimensions of different features and reduces the amount of computation. Ablation experiments demonstrate that this module increased success rates by 1.7% and precision by 0.5%. We evaluate CAFSN on challenging benchmarks such as OTB100, UAV123, and GOT-10K, validating advanced performance in noise immunity and similar target identification with 58.44 FPS in real time.


  • Open Access

    A Fast Circle Detector with Efficient Arc Extraction

    by Yang Liu, Honggui Deng, Zeyu Zhang, Qiguo Xu

    Circle detection is a crucial problem in computer vision and pattern recognition. Improving the accuracy and efficiency of circle detectors has important scientific significance and excellent application value. In this paper, we propose a circle detection method with efficient arc extraction. In order to reduce edge redundancy and eliminate crossing points, we present an edge refinement algorithm to refine the edges into single-pixel-wide branchless contour curves. To address the contour curve segmentation difficulty, we improved the CTAR (Chord to Triangular Arms Ratio) corner detection method to enhance corner point detection and segment the contour curves based on corner points. Then, we used the relative position constraint of arcs to improve the circle detection accuracy further. Finally, we verified the feasibility and reliability of the proposed method by comparing our approach with five other methods using three datasets. The experimental results showed that the presented method had the advantages of anti-obscuration, anti-defect, and real-time performance over other methods.


  • Open Access

    A Semi-Supervised Semantic Segmentation Method for Blast-Hole Detection

    by Zeyu Zhang, Honggui Deng, Yang Liu, Qiguo Xu, Gang Liu

    The goal of blast-hole detection is to help place charge explosives into blast-holes. This process is full of challenges, because it requires the ability to extract sample features in complex environments, and to detect a wide variety of blast-holes. Detection techniques based on deep learning with RGB-D semantic segmentation have emerged in recent years of research and achieved good results. However, implementing semantic segmentation based on deep learning usually requires a large amount of labeled data, which creates a large burden on the production of the dataset. To address the dilemma that there is very little training data available for explosive charging equipment to detect blast-holes, this paper extends the core idea of semi-supervised learning to RGB-D semantic segmentation, and devises an ERF-AC-PSPNet model based on a symmetric encoder–decoder structure. The model adds a residual connection layer and a dilated convolution layer for down-sampling, followed by an attention complementary module to acquire the feature maps, and uses a pyramid scene parsing network to achieve hole segmentation during decoding. A new semi-supervised learning method, based on pseudo-labeling and self-training, is proposed, to train the model for intelligent detection of blast-holes. The designed pseudo-labeling is based on the HOG algorithm and depth data, and proved to have good results in experiments. To verify the validity of the method, we carried out experiments on the images of blast-holes collected at a mine site. Compared to the previous segmentation methods, our method is less dependent on the labeled data and achieved IoU of 0.810, 0.867, 0.923, and 0.945, at labeling ratios of 1/8, 1/4, 1/2, and 1.


  • Open Access

    Efficient Task Planning for Mobile Manipulation: a Virtual Kinematic Chain Perspective

    by Ziyuan Jiao, Zeyu Zhang, Weiqi Wang, David Han, Song-Chun Zhu, Yixin Zhu, Hangxin Liu

    We present a Virtual Kinematic Chain (VKC) perspective, a simple yet effective method, to improve task planning efficacy for mobile manipulation. By consolidating the kinematics of the mobile base, the arm, and the object being manipulated collectively as a whole, this novel VKC perspective naturally defines abstract actions and eliminates unnecessary predicates in describing intermediate poses. As a result, these advantages simplify the design of the planning domain and significantly reduce the search space and branching factors in solving planning problems. In experiments, we implement a task planner using Planning Domain Definition Language (PDDL) with VKC. Compared with conventional domain definition, our VKC-based domain definition is more efficient in both planning time and memory. In addition, abstract actions perform better in producing feasible motion plans and trajectories. We further scale up the VKC-based task planner in complex mobile manipulation tasks. Taken together, these results demonstrate that task planning using VKC for mobile manipulation is not only natural and effective but also introduces new capabilities.