All 4 Comparisons

Input Image

Pretrained Stable Video Diffusion

Finetuned Stable Video Diffusion

Track4Gen without Refiner

Track4Gen

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Input Image

Limitation of all baselines (including Track4Gen)

Input Image

Pretrained Stable Video Diffusion

Finetuned Stable Video Diffusion

Track4Gen without Refiner

Track4Gen

Input Image

Input Image

Input Image

Input Image