-
15 Jan 2020: Our paper "Mastering Basketball with Deep Reinforcement Learning: An Integrated Curriculum Training Approach" and "Efficient Deep Reinforcement Learning through Policy Transfer" are accepted as extended abstract by AAMAS 2020.
-
20 Dec 2019: Our paper "Action Relation Network: Considering the Effects of Actions in Multiagent Systems" is accepted by ICLR 2020.
-
19 Dec 2019: Our presentation "Building Intelligent Game Testing System in Netease MMORPG Game" is accepted by AI Summit GDC 2020.
-
11 Nov 2019: Our 2 regular and 1 workshop papers are accepted by AAAI 2020, "Multi-Agent Game Abstraction via Graph Attention Neural Networks", "From Less to More: A Dynamic Agent Number Network for Large-scale Multiagent Curriculum Learning", "Generative Adversarial Imitation Learning from Failure Experience".
-
9 Nov 2019: Our presentation "Applying Reinforcement Learning to Develop Game AI in NetEase Game" is accepted by GDC 2020 Core
NetEase Fuxi Lab is considered as the professional game AI research division in China. Dr. Harry Shum, an NAE foreign member who is worldly-known for his research on artificial intelligence has joined us as our chief advisor. Fuxi is founded on the principle of bridging artificial intelligence and electronic games through transdisciplinary research. We use cutting-edge technologies to reshape the audiovisual experience of our gamers. In the meantime, the mass data and virtual environment provided by our game platform also enable us to boost the development of AI frontier research.
We focus on researching big data platform, user persona, reinforcement learning, computer vision & graphics, natural language processing, speech synthesis & music generation, and their potentiality in game industry. All of our team members have both professional education background (THU, PKU, ZJU, USTC, WHU, etc.) and strong practice aptitude.