Hello! I’m Qiwei Liang (梁棋炜), a undergraduate student majoring in Robotics Engineering at Shenzhen University, expected to graduate in July 2026. Currently, I am conducting research at the Humanoid Computing Lab, HKUST(gz), under the supervision of Renjing Xu.
Before diving into research, I was an IoT enthusiast dedicated to developing IoT products with the goal of improving people’s lives. I was also the national champion of the China Collegiate IoT Design Competition (物联网竞赛全国总冠军).
My research interests lie in Embodied AI, particularly in Robotic Manipulation (VLA, VA, RL), Robot Learning, and Whole-Body Control. Fundamentally, my work revolves around Exploring the Pinnacle of Wisdom.
First / Co-first Author25
- (ESWA 一区Top) STAR: Empowering Semi-Supervised Medical Image Segmentation with SAM-based Teacher-Student Architecture and Contrastive Consistency Regularization, paper
- (UR) Video2Reward++: Advancing Robot Skill Acquisition via Video-guided Reward Generation
- (CINT 2024) iKnowiSee: AR Glasses with Language Learning Translation System and Identity Recognition System Built Based on Large Pre-trained Models of Language and Vision and Internet of Things Technology, [paper]
- (arxiv 2025) Whole-Body Coordination for Dynamic Object Grasping with Legged Manipulators, [webpage]
Co-Author
- (ECAI 2024) Video2Reward: Generating Reward Function from Videos for Legged Robot Behavior Learning, [paper], [webpage]
- (arxiv 2025) RoboTwin 2.0: Autonomous Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation, [paper], [webpage]
- (arxiv 2025) FASR-Net: Unsupervised Shadow Removal Leveraging Inherent Frequency Priors, [paper]
- (ICLR WS 2025) Text2World: Benchmarking Large Language Models for Symbolic World Model Generation, [paper], [webpage]
- ✉️ kola337599@gmail.com
- 🌐 Personal site: https://kolakivy.github.io/
- 📚 Google Scholar / Github / Xiaohongshu links are on my website.
