Mengyang Wu
About
I am a research engineer at Huawei Hong Kong Research Center (HKRC) working on LLM/VLM/Agent Testing. Previously, I received my PhD degree from The Chinese University of Hong Kong (CUHK) and a Bachelor's degree from University College London (UCL). My research interests include 3D vision, scene understanding, and VLM/LLM-based agent applications.
Feel free to reach out if you'd like to collaborate or discuss research topics.
Publications
From Exploration to Exploitation: A Two-Stage Entropy RLVR Approach for Noise-Tolerant MLLM Training
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
VP-Bench: A Comprehensive Benchmark for Visual Prompting in Multimodal Large Language Models
AAAI Conference on Artificial Intelligence, 2026
KG-RAG: Enhancing GUI Agent Decision-Making via Knowledge Graph-Driven Retrieval-Augmented Generation
Empirical Methods in Natural Language Processing (EMNLP), 2025
Geometrically-plausible and Semantically-consistent Generation of Indoor Panoramas
IEEE International Conference on Multimedia and Expo (ICME), 2025
ICM-Assistant: Instruction-tuning Multimodal Large Language Models for Rule-based Explainable Image Content Moderation
AAAI Conference on Artificial Intelligence, 2025
Llava-spacesgg: Visual instruct tuning for open-vocabulary scene graph generation with enhanced spatial relations
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2025
Towards real-world adverse weather image restoration: Enhancing clearness and semantics with vision-language models
European Conference on Computer Vision (ECCV), 2024
FloorLevel-Net: recognizing floor-level lines with height-attention-guided multi-task learning
IEEE Transactions on Image Processing, 2021
Deep recognition of vanishing-point-constrained building planes in urban street views
IEEE Transactions on Image Processing, 2020