eLife

The view tolerance of human identity recognition depends on horizontal face information

View-tolerant identity recognition in humans depends on the stable identity cues provided by horizontal facial information.
Tech Times on MSN

VideoChat3 beats GPT-5 on video grounding: Open-source, full training stack released

VideoChat3, a 4-billion-parameter open-source video AI model from Nanjing University, outperforms GPT-5 and Gemini 2.5 Flash ...