arXiv subjects

Man Zhang

Publications and source records attributed to Man Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2026

Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Models

Text-to-image (T2I) models have achieved remarkable success in generating high-fidelity images, but they often fail in handling complex spatial relationships, e.g., spatial perception, reasoning, or interaction. These critical aspects are largely overlooked by current benchmarks due to their short or information-sparse prompt design. In this paper, we introduce SpatialGenEval, a new benchmark designed to systematically evaluate the spatial intelligence of T2I models, covering two key aspects: (1) SpatialGenEval involves 1,230 long, information-dense prompts across 25 real-world scenes. Each prompt integrates 10 spatial sub-domains and corresponding 10 multi-choice question-answer pairs, ranging from object position and layout to occlusion and causality. Our extensive evaluation of 21 state-of-the-art models reveals that higher-order spatial reasoning remains a primary bottleneck. (2) To demonstrate that the utility of our information-dense design goes beyond simple evaluation, we also construct the SpatialT2I dataset. It contains 15,400 text-image pairs with rewritten prompts to ensure image consistency while preserving information density. Fine-tuned results on current foundation models (i.e., Stable Diffusion-XL, Uniworld-V1, OmniGen2) yield consistent performance gains (+4.2%, +5.7%, +4.4%) and more realistic effects in spatial relations, highlighting a data-centric paradigm to achieve spatial intelligence in T2I models.

cs.CV↗

arXiv2026

Learning Geometric Invariance for Gait Recognition

The goal of gait recognition is to extract identity-invariant features of an individual under various gait conditions, e.g., cross-view and cross-clothing. Most gait models strive to implicitly learn the common traits across different gait conditions in a data-driven manner to pull different gait conditions closer for recognition. However, relatively few studies have explicitly explored the inherent relations between different gait conditions. For this purpose, we attempt to establish connections among different gait conditions and propose a new perspective to achieve gait recognition: variations in different gait conditions can be approximately viewed as a combination of geometric transformations. In this case, all we need is to determine the types of geometric transformations and achieve geometric invariance, then identity invariance naturally follows. As an initial attempt, we explore three common geometric transformations (i.e., Reflect, Rotate, and Scale) and design a $\mathcal{R}$eflect-$\mathcal{R}$otate-$\mathcal{S}$cale invariance learning framework, named ${\mathcal{RRS}}$-Gait. Specifically, it first flexibly adjusts the convolution kernel based on the specific geometric transformations to achieve approximate feature equivariance. Then these three equivariant-aware features are respectively fed into a global pooling operation for final invariance-aware learning. Extensive experiments on four popular gait datasets (Gait3D, GREW, CCPG, SUSTech1K) show superior performance across various gait conditions.

cs.CV↗

arXiv2025

RoleMotion: A Large-Scale Dataset towards Robust Scene-Specific Role-Playing Motion Synthesis with Fine-grained Descriptions

In this paper, we introduce RoleMotion, a large-scale human motion dataset that encompasses a wealth of role-playing and functional motion data tailored to fit various specific scenes. Existing text datasets are mainly constructed decentrally as amalgamation of assorted subsets that their data are nonfunctional and isolated to work together to cover social activities in various scenes. Also, the quality of motion data is inconsistent, and textual annotation lacks fine-grained details in these datasets. In contrast, RoleMotion is meticulously designed and collected with a particular focus on scenes and roles. The dataset features 25 classic scenes, 110 functional roles, over 500 behaviors, and 10296 high-quality human motion sequences of body and hands, annotated with 27831 fine-grained text descriptions. We build an evaluator stronger than existing counterparts, prove its reliability, and evaluate various text-to-motion methods on our dataset. Finally, we explore the interplay of motion generation of body and hands. Experimental results demonstrate the high-quality and functionality of our dataset on text-driven whole-body generation.

cs.CV↗

arXiv2025

SIG-Chat: Spatial Intent-Guided Conversational Gesture Generation Involving How, When and Where

The accompanying actions and gestures in dialogue are often closely linked to interactions with the environment, such as looking toward the interlocutor or using gestures to point to the described target at appropriate moments. Speech and semantics guide the production of gestures by determining their timing (WHEN) and style (HOW), while the spatial locations of interactive objects dictate their directional execution (WHERE). Existing approaches either rely solely on descriptive language to generate motions or utilize audio to produce non-interactive gestures, thereby lacking the characterization of interactive timing and spatial intent. This significantly limits the applicability of conversational gesture generation, whether in robotics or in the fields of game and animation production. To address this gap, we present a full-stack solution. We first established a unique data collection method to simultaneously capture high-precision human motion and spatial intent. We then developed a generation model driven by audio, language, and spatial data, alongside dedicated metrics for evaluating interaction timing and spatial accuracy. Finally, we deployed the solution on a humanoid robot, enabling rich, context-aware physical interactions.

cs.GR↗

arXiv2025

Empirical Studies on Quantum Optimization for Software Engineering: A Systematic Analysis

In recent years, quantum, quantum-inspired, and hybrid algorithms are increasingly showing promise for solving software engineering optimization problems. However, best-intended practices for conducting empirical studies have not yet well established. In this paper, based on the primary studies identified from the latest systematic literature review on quantum optimization for software engineering problems, we conducted a systematic analysis on these studies from various aspects including experimental designs, hyperparameter settings, case studies, baselines, tooling, and metrics. We identify key gaps in the current practices such as limited reporting of the number of repetitions, number of shots, and inadequate consideration of noise handling, as well as a lack of standardized evaluation protocols such as the adoption of quality metrics, especially quantum-specific metrics. Based on our analysis, we provide insights for designing empirical studies and highlight the need for more real-world and open case studies to assess cost-effectiveness and practical utility of the three types of approaches: quantum-inspired, quantum, and hybrid. This study is intended to offer an overview of current practices and serve as an initial reference for designing and conducting empirical studies on evaluating and comparing quantum, quantum-inspired, and hybrid algorithms in solving optimization problems in software engineering.

cs.SE↗

arXiv2025

WFC/WFD: Web Fuzzing Commons, Dataset and Guidelines to Support Experimentation in REST API Fuzzing

Fuzzing REST APIs is an important research problem, with practical applications and impact in industry. As such, a lot of research work has been carried out on this topic in the last few years. However, there are three major issues that hinder further progress: how to deal with API authentication; how to catalog and compare different fault types found by different fuzzers; and what to use as case study to facilitate fair comparisons among fuzzers. To address these important challenges, we present Web Fuzzing Commons (WFC) and Web Fuzzing Dataset (WFD). WFC is a set of open-source libraries and schema definitions to declaratively specify authentication info and catalog different types of faults that fuzzers can automatically detect. WFD is a collection of 36 open-source APIs with all necessary scaffolding to easily run experiments with fuzzers, supported by WFC. To show the usefulness of WFC/WFD, a set of experiments is carried out with EvoMaster, a state-of-the-art fuzzer for Web APIs. However, any fuzzer can benefit from WFC and WFD. We compare EvoMaster with other state-of-the-art tools such as ARAT-RL, EmRest, LLamaRestTest, RESTler, and Schemathesis. We discuss common pitfalls in tool comparisons, as well as providing guidelines with support of WFC/WFD to avoid them.

cs.SE↗

arXiv2025

DanceEditor: Towards Iterative Editable Music-driven Dance Generation with Open-Vocabulary Descriptions

Generating coherent and diverse human dances from music signals has gained tremendous progress in animating virtual avatars. While existing methods support direct dance synthesis, they fail to recognize that enabling users to edit dance movements is far more practical in real-world choreography scenarios. Moreover, the lack of high-quality dance datasets incorporating iterative editing also limits addressing this challenge. To achieve this goal, we first construct DanceRemix, a large-scale multi-turn editable dance dataset comprising the prompt featuring over 25.3M dance frames and 84.5K pairs. In addition, we propose a novel framework for iterative and editable dance generation coherently aligned with given music signals, namely DanceEditor. Considering the dance motion should be both musical rhythmic and enable iterative editing by user descriptions, our framework is built upon a prediction-then-editing paradigm unifying multi-modal conditions. At the initial prediction stage, our framework improves the authority of generated results by directly modeling dance movements from tailored, aligned music. Moreover, at the subsequent iterative editing stages, we incorporate text descriptions as conditioning information to draw the editable results through a specifically designed Cross-modality Editing Module (CEM). Specifically, CEM adaptively integrates the initial prediction with music and text prompts as temporal motion cues to guide the synthesized sequences. Thereby, the results display music harmonics while preserving fine-grained semantic alignment with text descriptions. Extensive experiments demonstrate that our method outperforms the state-of-the-art models on our newly collected DanceRemix dataset. Code is available at https://lzvsdy.github.io/DanceEditor/.

cs.GR↗

arXiv2025

GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

We present GLM-4.5, an open-source Mixture-of-Experts (MoE) large language model with 355B total parameters and 32B activated parameters, featuring a hybrid reasoning method that supports both thinking and direct response modes. Through multi-stage training on 23T tokens and comprehensive post-training with expert model iteration and reinforcement learning, GLM-4.5 achieves strong performance across agentic, reasoning, and coding (ARC) tasks, scoring 70.1% on TAU-Bench, 91.0% on AIME 24, and 64.2% on SWE-bench Verified. With much fewer parameters than several competitors, GLM-4.5 ranks 3rd overall among all evaluated models and 2nd on agentic benchmarks. We release both GLM-4.5 (355B parameters) and a compact version, GLM-4.5-Air (106B parameters), to advance research in reasoning and agentic AI systems. Code, models, and more information are available at https://github.com/zai-org/GLM-4.5.

cs.CL↗

arXiv2025

Search-Based Fuzzing For RESTful APIs That Use MongoDB

In RESTful APIs, interactions with a database are a common and crucial aspect. When generating whitebox tests, it is essential to consider the database's state (i.e., the data contained in the database) to achieve higher code coverage and uncover more hidden faults. This article presents novel techniques to enhance search-based software test generation for RESTful APIs interacting with NoSQL databases. Specifically, we target the popular MongoDB database, by dynamically analyzing (via automated code instrumentation) the state of the database during the test generation process. Additionally, to achieve better results, our novel approach allows inserting NoSQL data directly from test cases. This is particularly beneficial when generating the correct sequence of events to set the NoSQL database in an appropriate state is challenging or time-consuming. This method is also advantageous for testing read-only microservices. Our novel techniques are implemented as an extension of EvoMaster, the only open-source tool for white-box fuzzing RESTful APIs. Experiments conducted on six RESTful APIs demonstrated significant improvements in code coverage, with increases of up to 18% compared to existing white-box approaches. To better highlight the improvements of our novel techniques, comparisons are also carried out with four state-of-the-art black-box fuzzers.

cs.SE↗

arXiv2025

ReMeREC: Relation-aware and Multi-entity Referring Expression Comprehension

Referring Expression Comprehension (REC) aims to localize specified entities or regions in an image based on natural language descriptions. While existing methods handle single-entity localization, they often ignore complex inter-entity relationships in multi-entity scenes, limiting their accuracy and reliability. Additionally, the lack of high-quality datasets with fine-grained, paired image-text-relation annotations hinders further progress. To address this challenge, we first construct a relation-aware, multi-entity REC dataset called ReMeX, which includes detailed relationship and textual annotations. We then propose ReMeREC, a novel framework that jointly leverages visual and textual cues to localize multiple entities while modeling their inter-relations. To address the semantic ambiguity caused by implicit entity boundaries in language, we introduce the Text-adaptive Multi-entity Perceptron (TMP), which dynamically infers both the quantity and span of entities from fine-grained textual cues, producing distinctive representations. Additionally, our Entity Inter-relationship Reasoner (EIR) enhances relational reasoning and global scene understanding. To further improve language comprehension for fine-grained prompts, we also construct a small-scale auxiliary dataset, EntityText, generated using large language models. Experiments on four benchmark datasets show that ReMeREC achieves state-of-the-art performance in multi-entity grounding and relation prediction, outperforming existing approaches by a large margin.

cs.CV↗

arXiv2025

Quantum Optimization for Software Engineering: A Survey

Quantum computing, particularly in the area of quantum optimization, is steadily progressing toward practical applications, supported by an expanding range of hardware platforms and simulators. While Software Engineering (SE) optimization has a strong foundation, which is exemplified by the active Search-Based Software Engineering (SBSE) community and numerous classical optimization methods, the growing complexity of modern software systems and their engineering processes demands innovative solutions. This Systematic Literature Review (SLR) focuses specifically on studying the literature that applies quantum or quantum-inspired algorithms to solve classical SE optimization problems. We examine 77 primary studies selected from an initial pool of 2083 publications obtained through systematic searches of six digital databases using carefully crafted search strings. Our findings reveal concentrated research efforts in areas such as SE operations and software testing, while exposing significant gaps across other SE activities. Additionally, the SLR uncovers relevant works published outside traditional SE venues, underscoring the necessity of this comprehensive review. Overall, our study provides a broad overview of the research landscape, empowering the SBSE community to leverage quantum advancements in addressing next-generation SE challenges.

cs.SE↗

arXiv2025

VividListener: Expressive and Controllable Listener Dynamics Modeling for Multi-Modal Responsive Interaction

Generating responsive listener head dynamics with nuanced emotions and expressive reactions is crucial for practical dialogue modeling in various virtual avatar animations. Previous studies mainly focus on the direct short-term production of listener behavior. They overlook the fine-grained control over motion variations and emotional intensity, especially in long-sequence modeling. Moreover, the lack of long-term and large-scale paired speaker-listener corpora including head dynamics and fine-grained multi-modality annotations (e.g., text-based expression descriptions, emotional intensity) also limits the application of dialogue modeling.Therefore, we first newly collect a large-scale multi-turn dataset of 3D dyadic conversation containing more than 1.4M valid frames for multi-modal responsive interaction, dubbed ListenerX. Additionally, we propose VividListener, a novel framework enabling fine-grained, expressive and controllable listener dynamics modeling. This framework leverages multi-modal conditions as guiding principles for fostering coherent interactions between speakers and listeners.Specifically, we design the Responsive Interaction Module (RIM) to adaptively represent the multi-modal interactive embeddings. RIM ensures the listener dynamics achieve fine-grained semantic coordination with textual descriptions and adjustments, while preserving expressive reaction with speaker behavior. Meanwhile, we design the Emotional Intensity Tags (EIT) for emotion intensity editing with multi-modal information integration, applying to both text descriptions and listener motion amplitude.Extensive experiments conducted on our newly collected ListenerX dataset demonstrate that VividListener achieves state-of-the-art performance, realizing expressive and controllable listener dynamics.

cs.CV↗

arXiv2025

Advances in Embodied Navigation Using Large Language Models: A Survey

In recent years, the rapid advancement of Large Language Models (LLMs) such as the Generative Pre-trained Transformer (GPT) has attracted increasing attention due to their potential in a variety of practical applications. The application of LLMs with Embodied Intelligence has emerged as a significant area of focus. Among the myriad applications of LLMs, navigation tasks are particularly noteworthy because they demand a deep understanding of the environment and quick, accurate decision-making. LLMs can augment embodied intelligence systems with sophisticated environmental perception and decision-making support, leveraging their robust language and image-processing capabilities. This article offers an exhaustive summary of the symbiosis between LLMs and embodied intelligence with a focus on navigation. It reviews state-of-the-art models, research methodologies, and assesses the advantages and disadvantages of existing embodied navigation models and datasets. Finally, the article elucidates the role of LLMs in embodied intelligence, based on current research, and forecasts future directions in the field. A comprehensive list of studies in this survey is available at https://github.com/Rongtao-Xu/Awesome-LLM-EN.

cs.AI↗

arXiv2025

AI Delivers Creative Output but Struggles with Thinking Processes

A key objective in artificial intelligence (AI) development is to create systems that match or surpass human creativity. Although current AI models perform well across diverse creative tasks, it remains unclear whether these achievements reflect genuine creative thinking. This study examined whether AI models (GPT-3.5-turbo, GPT-4, and GPT-4o) engage in creative thinking by comparing their performance with humans across various creative tasks and core cognitive processes. Results showed that AI models outperformed humans in divergent thinking, convergent thinking, and insight problem-solving, but underperformed in creative writing. Compared to humans, AI generated lower forward flow values in both free and chain association tasks and showed lower accuracy in the representational change task. In creative evaluation, AI exhibited no significant correlation between the weights of novelty and appropriateness when predicting creative ratings, suggesting the absence of a human-like trade-off strategy. AI also had higher decision error scores in creative selection, suggesting difficulty identifying the most creative ideas. These findings suggest that while AI can mimic human creativity, its strong performance in creative tasks is likely driven by non-creative mechanisms rather than genuine creative thinking.

cs.HC↗

arXiv2025

SIP-IFVM: A time-evolving coronal model with an extended magnetic field decomposition strategy

Time-evolving magnetohydrodynamic (MHD) coronal modeling, driven by a series of time-dependent photospheric magnetograms, represents a new generation of coronal simulations. This approach offers greater realism compared to traditional coronal models constrained by a static magnetogram. However, its practical application is seriously limited by low computational efficiency and poor numerical stability. Therefore, we propose an extended magnetic field decomposition strategy and implement it in the implicit MHD model to develop a coronal model that is both efficient and numerically stable enough for simulating the long-term evolutions of the global corona. The traditional decomposition strategies split the magnetic field into a time-invariant potential field and a time-dependent component $\mathbf{B}_1$. It works well for quasi-steady-state coronal simulations where $\left|\mathbf{B}_1\right|$ is typically small. However, as the inner-boundary magnetic field evolves, $\left|\mathbf{B}_1\right|$ can grow significantly larger and its discretization errors often lead to nonphysical negative thermal pressure, ultimately causing the code to crash. In this paper, we mitigate such undesired situations by introducing a temporally piecewise-constant variable to accommodate part of the non-potential field and remain $\left|\mathbf{B}_1\right|$ consistently small throughout the simulations. We incorporate this novel magnetic field decomposition strategy into our implicit MHD coronal model and apply it to simulate the evolution of coronal structures within 0.1 AU over two solar-maximum Carrington rotations. The results show that this coronal model effectively captures observations and performs more than 80 times faster than real time using only 192 CPU cores, making it well-suited for practical applications in simulating the time-evolving corona.

astro-ph.SR↗

arXiv2025

Causes and Effects of Fitness Landscapes in System Test Generation: A Replication Study

Search-Based Software Testing (SBST) has seen several success stories in academia and industry. The effectiveness of a search algorithm at solving a software engineering problem strongly depends on how such algorithm can navigate the fitness landscape of the addressed problem. The fitness landscape depends on the used fitness function. Understanding the properties of a fitness landscape can help to provide insight on how a search algorithm behaves on it. Such insight can provide valuable information to researchers to being able to design novel, more effective search algorithms and fitness functions tailored for a specific problem. Due to its importance, few fitness landscape analyses have been carried out in the scientific literature of SBST. However, those have been focusing on the problem of unit test generation, e.g., with state-of-the-art tools such as EvoSuite. In this paper, we replicate one such existing study. However, in our work we focus on system test generation, with the state-of-the-art tool EvoMaster. Based on an empirical study involving the testing of 23 web services, this enables us to provide valuable insight into this important testing domain of practical industrial relevance.

cs.SE↗

arXiv2024

StableMoFusion: Towards Robust and Efficient Diffusion-based Motion Generation Framework

Thanks to the powerful generative capacity of diffusion models, recent years have witnessed rapid progress in human motion generation. Existing diffusion-based methods employ disparate network architectures and training strategies. The effect of the design of each component is still unclear. In addition, the iterative denoising process consumes considerable computational overhead, which is prohibitive for real-time scenarios such as virtual characters and humanoid robots. For this reason, we first conduct a comprehensive investigation into network architectures, training strategies, and inference processs. Based on the profound analysis, we tailor each component for efficient high-quality human motion generation. Despite the promising performance, the tailored model still suffers from foot skating which is an ubiquitous issue in diffusion-based solutions. To eliminate footskate, we identify foot-ground contact and correct foot motions along the denoising process. By organically combining these well-designed components together, we present StableMoFusion, a robust and efficient framework for human motion generation. Extensive experimental results show that our StableMoFusion performs favorably against current state-of-the-art methods. Project page: https://h-y1heng.github.io/StableMoFusion-page/

cs.CV↗

arXiv2024

Interplanetary Rotation of 2021 December 4 CME

The magnetic orientation of coronal mass ejections (CMEs) is of great importance to understand their space weather effects. Although many evidences suggest that CMEs can undergo significant rotation during the early phases of evolution in the solar corona, there are few reports that CMEs rotate in the interplanetary space. In this work, we use multi-spacecraft observations and a numerical simulation starting from the lower corona close to the solar surface to understand the CME event on 2021 December 4, with an emphatic investigation of its rotation. This event is observed as a partial halo CME from the back side of the Sun by coronagraphs, and reaches the BepiColombo spacecraft and the MAVEN/Tianwen-1 as a magnetic flux rope-like structure. The simulation discloses that in the solar corona the CME is approximately a translational motion, while the interplanetary propagation process evidences a gradual change of axis orientation of the CME's flux rope-like structure. It is also found that the downside and the right flank of the CME moves with the fast solar wind, and the upside does in the slow-speed stream. The different parts of the CME with different speeds generate the nonidentical displacements of its magnetic structure, resulting in the rotation of the CME in the interplanetary space. Furthermore, at the right flank of the CME exists a corotating interaction region (CIR), which makes the orientation of the CME alter, and also deviates from its route due to the CME. These results provide new insight on interpreting CMEs' dynamics and structures during their travelling through the heliosphere.

astro-ph.SR↗