Self Gradient Forcing: Native Long Video Extrapolation
- Published
- Source
- arXiv
- Paper number
- 690
- Field
- Computer Vision
- arXiv ID
- 2607.20368
Key points
- It defines and resolves the past-context gradient gap, which is the critical weakness of Self Forcing.
- A two-pass strategy, where the first pass records rollouts and the second pass reconstructs gradients in parallel, solves the problem without extra memory overhead.
- Using only 5-second training videos, it can generate coherent long videos up to 240 seconds, improving subject identity, background consistency, and temporal stability.
- A user study found that it was preferred over Self Forcing in every setting, with GSB ranging from 29.6% to 48.7%.
- Peak memory increases by only 8 GB, and training slowdown is about 13%, which makes the method practical.
Paper links
External research summaries. These are not HDATF publications or measured product results.