1 paper
Yanliang Qi, Kexi Chen, Muchao Ye +1
Text-conditioned image-to-video (I2V) generation has advanced rapidly, yet generating videos with multiple subjects remains challenging. A model must simultaneously preserve the ap…