Huang Tiejun

Huang earned bachelor’s and master’s degrees at Wuhan University of Technology and a doctorate in pattern recognition and intelligent systems at Huazhong University of Science and Technology. Peking University identifies him as a professor and vice dean of its Institute for Artificial Intelligence. BAAI’s 2026 conference release identifies him as chairman of its board.[1][5]

A toolkit for continuous visual signals

SpikeCV provides data, algorithms and hardware interfaces for spike-camera vision. These sensors accumulate light intensity and emit asynchronous spikes rather than periodic image frames, creating a need for different reconstruction and analysis tools. Huang is a paper coauthor and is credited by the project as its initiator and the spike-camera inventor. The framework is a team effort supporting experiments with that representation.[2][3]

One sequence for several media

Huang coauthored Emu3, which tokenizes text, images and video into discrete representations and trains a transformer to predict the next token. This brings generation and perception into one sequence-modeling formulation. The paper compares the resulting models with task-specific systems and releases techniques and models for further study. It presents a training approach for multimodal capabilities, with performance tested on the reported tasks.[4]