Research
Hugging Face Papers lists FAPL, a method for cinematic video generation using MLLM latents
FAPL combines latent-only generator grounding, factor-span replacement and a two-stage training scheme to improve sparse-to-video, factor-aware editing and multi-shot generation over baselines, according to the authors.
Every outlet that ran it 1 article
Hugging Face PapersResearchLead
Making MLLM Latents Matter: Factor-Aware Predictive Learning for Cinematic Video Generation