1 paper · 1 filter
Xiaoyang Han, Jianhua Li, Kewang Deng +14
We formulate computer vision as unified multimodal generation, where heterogeneous visual tasks are expressed in the native text and image generation spaces of a unified multimodal…