TY  - RPRT
TI  - Direct, Parallel, or Sequential? A Comparative Study of Training-Free Multi-Subject Image-to-Video Generation
AU  - Yanliang Qi
AU  - Kexi Chen
AU  - Muchao Ye
AU  - Haomiao Ni
PY  - 2026
DO  - 10.1145/3841458.3841542
UR  - https://arxiv.org/abs/2608.22819
ID  - 2608.22819
ER  -