SenseTime, the Chinese artificial intelligence company under US sanctions, has unveiled a new open-source model designed to process images faster than leading competitors. The model, called SenseNova U1, can generate and interpret images without first converting them to text, significantly reducing processing time and computational demand.
Released on Tuesday, SenseNova U1 marks a strategic pivot for SenseTime, which has struggled to maintain its position among China's top AI developers amid restricted access to advanced technology. By open-sourcing the model, the company aims to regain influence and demonstrate resilience.
"The model's entire reasoning process is no longer limited to text. It can reason with images as well," said Dahua Lin, SenseTime's cofounder and chief scientist, in an interview. Lin, also a professor at the Chinese University of Hong Kong, explained that image-direct processing could help robots better understand the physical world in the future.
SenseTime's focus on speed and efficiency comes as it relies on domestically produced chips, a necessity due to US export controls. The company hopes SenseNova U1's performance advantage will attract developers and rebuild its reputation in the competitive AI landscape.