arXiv subjects

Zihang Zeng

Publications and source records attributed to Zihang Zeng.

7 recordsLinked to original sources

arXiv2026

Low-Altitude Channel Multipath Prediction via Panoramic Perception and Vision-Language Model

Unmanned aerial vehicle (UAV) communication is expected to support a wide range of low-altitude applications in 6G mobile networks. However, traditional statistical channel models provide limited accuracy in specific environments, while deterministic methods such as ray tracing usually rely on accurate three-dimensional environment models and involve high computational complexity. Existing multimodal channel prediction approaches mainly focus on large-scale metrics such as path loss, and remain insufficient for modeling small-scale parameters. To address these limitations, this paper proposes PanoLAMP, a Panoramic perception and vision-language model-based Low-Altitude Multipath Prediction framework. It adopts a pretrained vision-language model as the backbone and captures the propagation environment features through panoramic RGB-D observations collected at both the transmitter and receiver to predict the delay, power, azimuth angle, and zenith angle offset relative to the line-of-sight path. Experiments are conducted on a synthetic dataset containing 18,949 UAV-vehicle links across seven UAV altitudes. Experimental results show that the proposed method consistently outperforms representative baselines in both multipath parameters and statistical metrics, and demonstrates stronger generalization across different flight heights.

cs.CV↗

arXiv2026

Video2Sim2Real: Full-Stack Autonomous Dexterous Skill Acquisition from a Single Human Video

Human manipulation videos are a convenient and intuitive source for robot learning. However, directly transferring human dexterity to robots remains challenging due to perception errors and embodiment gap. To address this, we introduce Video2Sim2Real, a full-stack framework for autonomous skill acquisition from a single human manipulation video. Our framework first uses off-the-shelf foundation models to reconstruct a simulator-ready digital twin and extract robot and object motion priors. Rather than treating the extracted robot motion as a reliable reference throughout execution, our key idea is to recover and leverage the most fundamental sources of supervision from the demonstrated skill: We identify object-centric keyframes to optimize the corresponding robot configurations using object information from the simulator, and use these configurations as anchors that refine the robot motion such that it ultimately has the desired impact on the environment. To bridge the remaining sim-to-real gap, we introduce a sim-to-real strategy that decouples robustness to noisy and incomplete perception from variations in hand-object interaction dynamics. Specifically, we learn to recalibrate robot configurations from noisy real-world point clouds via IL, and leverage residual RL to perform local finger-level adaptations to ensure for robust and effective interactions. Finally, a collision-aware motion planning module enables spatial generalization to novel object configurations. Across several everyday manipulation tasks, Video2Sim2Real improves simulated task success, safety, and trajectory coherence over numerous baselines, and achieves better sim-to-real transfer than existing techniques. These results demonstrate a promising path toward autonomous dexterous skill acquisition from human videos.

cs.RO↗

arXiv2026

Hidden in Plain Sight: Visual-to-Symbolic Analytical Solution Inference from Field Visualizations

Recovering analytical solutions of physical fields from visual observations is a fundamental yet underexplored capability for AI-assisted scientific reasoning. We study visual-to-symbolic analytical solution inference (ViSA) for two-dimensional linear steady-state fields: given field visualizations (and first-order derivatives) plus minimal auxiliary metadata, the model must output a single executable SymPy expression with fully instantiated numeric constants. We introduce ViSA-R2 and align it with a self-verifying, solution-centric chain-of-thought pipeline that follows a physicist-like pathway: structural pattern recognition solution-family (ansatz) hypothesis parameter derivation consistency verification. We also release ViSA-Bench, a VLM-ready synthetic benchmark covering 30 linear steady-state scenarios with verifiable analytical/symbolic annotations, and evaluate predictions by numerical accuracy, expression-structure similarity, and character-level accuracy. Using an 8B open-weight Qwen3-VL backbone, ViSA-R2 outperforms strong open-source baselines and the evaluated closed-source frontier VLMs under a standardized protocol.

cs.AI↗

arXiv2026

AI-for-Science Low-code Platform with Bayesian Adversarial Multi-Agent Framework

Large Language Models (LLMs) demonstrate potentials for automating scientific code generation but face challenges in reliability, error propagation in multi-agent workflows, and evaluation in domains with ill-defined success metrics. We present a Bayesian adversarial multi-agent framework specifically designed for AI for Science (AI4S) tasks in the form of a Low-code Platform (LCP). Three LLM-based agents are coordinated under the Bayesian framework: a Task Manager that structures user inputs into actionable plans and adaptive test cases, a Code Generator that produces candidate solutions, and an Evaluator providing comprehensive feedback. The framework employs an adversarial loop where the Task Manager iteratively refines test cases to challenge the Code Generator, while prompt distributions are dynamically updated using Bayesian principles by integrating code quality metrics: functional correctness, structural alignment, and static analysis. This co-optimization of tests and code reduces dependence on LLM reliability and addresses evaluation uncertainty inherent to scientific tasks. LCP also streamlines human-AI collaboration by translating non-expert prompts into domain-specific requirements, bypassing the need for manual prompt engineering by practitioners without coding backgrounds. Benchmark evaluations demonstrate LCP's effectiveness in generating robust code while minimizing error propagation. The proposed platform is also tested on an Earth Science cross-disciplinary task and demonstrates strong reliability, outperforming competing models.

cs.AI↗

arXiv2025

VoxelRF: Voxelized Radiance Field for Fast Wireless Channel Modeling

Wireless channel modeling in complex environments is crucial for wireless communication system design and deployment. Traditional channel modeling approaches face challenges in balancing accuracy, efficiency, and scalability, while recent neural approaches such as neural radiance field (NeRF) suffer from long training and slow inference. To tackle these challenges, we propose voxelized radiance field (VoxelRF), a novel neural representation for wireless channel modeling that enables fast and accurate synthesis of spatial spectra. VoxelRF replaces the costly multilayer perception (MLP) used in NeRF-based methods with trilinear interpolation of voxel grid-based representation, and two shallow MLPs to model both propagation and transmitter-dependent effects. To further accelerate training and improve generalization, we introduce progressive learning, empty space skipping, and an additional background entropy loss function. Experimental results demonstrate that VoxelRF achieves competitive accuracy with significantly reduced computation and limited training data, making it more practical for real-time and resource-constrained wireless applications.

eess.SP↗

arXiv2022

FedNLP: Benchmarking Federated Learning Methods for Natural Language Processing Tasks

Increasing concerns and regulations about data privacy and sparsity necessitate the study of privacy-preserving, decentralized learning methods for natural language processing (NLP) tasks. Federated learning (FL) provides promising approaches for a large number of clients (e.g., personal devices or organizations) to collaboratively learn a shared global model to benefit all clients while allowing users to keep their data locally. Despite interest in studying FL methods for NLP tasks, a systematic comparison and analysis is lacking in the literature. Herein, we present the FedNLP, a benchmarking framework for evaluating federated learning methods on four different task formulations: text classification, sequence tagging, question answering, and seq2seq. We propose a universal interface between Transformer-based language models (e.g., BERT, BART) and FL methods (e.g., FedAvg, FedOPT, etc.) under various non-IID partitioning strategies. Our extensive experiments with FedNLP provide empirical comparisons between FL methods and helps us better understand the inherent challenges of this direction. The comprehensive analysis points to intriguing and exciting future research aimed at developing FL methods for NLP tasks.

cs.CL↗

arXiv2019

Tumor Microenvironment-based Gene Signatures Divides Novel Immune and Stromal Subgroup Classification of Lung Adenocarcinoma

Tumor microenvironment has complex effects on tumorigenesis and metastasis. However, there is still a lack of comprehensive understanding of the relationship among molecular and cellular characteristics in tumor microenvironment, clinical prognosis and immunotherpy response. In this study, the immune and stromal (non-immune) signatures of tumor microenvironment were integrated to identify novel subgroups of lung adenocarcinoma by eigendecomposition and extraction algorithms of bioinformatics and machine learning, such as non-negative matrix factorization and multitask learning. Tumors were classified into 4 groups according to the activation of immunity and stroma by novel signatures. The 4 groups had different mutation landscape, molecular, cellular characteristics and prognosis, which have been validation in 6 independent data sets containing 1551 patients. High-immune and low-stromal activation group links to high immunocyte infiltration, high immunocompetence, low fibroblasts, endothelial cells, collagen, laminin, tumor mutation burden, and better overall survival. We developed a novel model based on tumor microenvironment by integrating immune and stromal activation, namely PMBT (prognostic model based on tumor microenvironment). The PMBT showed the value to predict overall survival and immunotherapy responses.

q-bio.GN↗