Looking for a job

Hongyang Xue

@hongyangxue

Message

AI researcher at A5 Labs (AceGuardian) researching poker anti-cheating and policies that guide opponent selection.

China

Message

What I'm looking for

I'm interested in AI research, especially exploratory research that develops new ideas from the ground up.

At A5 Labs (AceGuardian), I research strategy-level anti-cheating algorithms for poker, including in-context learning. This work helped identify bot types that traditional anti-cheating methods missed.

I also researched policy representation learning; the resulting models and algorithms are now used by the Gameplay team for opponent selection.

At Yiwise, I developed digital human algorithms for short-video production and live streaming, including real-time audio-to-face and one-shot video face swapping. I also developed a green-screen keying algorithm using OpenGL and GLSL shaders, which reached 1,000 fps on an RTX 3060 GPU.

At Inspir.AI, I conducted exploratory research in imitation learning and placed third in the MineRL 2021 competition. Earlier, at NetEase Games AI Lab, I developed a distributed reinforcement-learning training framework and ranked fourth in the Google Research Football 11-vs-11 competition.

Experience

Work history, roles, and key accomplishments

Current

Senior AI Researcher

Current

A5 Labs

Jul 2025 - Present (1 year 3 months)

Research strategy-level anti-cheating algorithms for poker, exploring techniques such as in-context learning. Research models and algorithms for policy representation learning, now used by the Gameplay team for opponent selection.

Game AI Anti Cheat Development In Context Learning Policy Representation Opponent Selection

Technical Partner, Head of AI Lab

Yiwise

Dec 2022 - Jul 2025 (2 years 7 months)

Developed a 2D audio-to-face algorithm supporting real-time streaming audio in multiple languages, with highly consistent audio and lip movements. Developed a high-performance green-screen keying algorithm based on traditional image processing, implemented with OpenGL and GLSL shaders, achieving 1,000 fps on an RTX 3060 GPU.

Digital Humans GenAI Agents Loop Engineering Lip Sync Face Swap

Algorithm Tech Lead

Inspir.AI

Dec 2020 - Dec 2022 (1 year 11 months)

Conducted exploratory research in imitation learning and placed third in the MineRL 2021 competition. Designed neural network architectures, including Transformers, CNNs, and LSTMs, for multimodal game observations such as vectors, images, and sets.

Game AI Reinforcement Learning Imitation Learning Multi Technology Networks Hierarchical Policies

Senior AI Research Engineer

NetEase Games AI Lab

Jul 2019 - Dec 2020 (1 year 5 months)

Conducted exploratory research in reinforcement learning and imitation learning, including DQN variants, PPO, SAC, IMPALA, and GAIL. Developed a distributed RL training framework supporting PPO, DDQN, and SAC, with self-play, fictitious self-play, and prioritized self-play.

Multi Agent RL PPO Imitation Learning Self Play League Training

Education

Degrees, certifications, and relevant coursework

Zhejiang University

PhD, Computer Science

2014 - 2019

PhD in Computer Science at Zhejiang University, supervised by Deng Cai and Xiaofei He, with research in video question answering, stereo matching, and image processing.