1 paper
Yang-Tian Sun, Zehuan Huang, Yifan Niu +4
We present StereoWorld, a camera-conditioned stereo world model that jointly learns appearance and binocular geometry for end-to-end stereo video generation.Unlike monocular RGB or…