Search papers, labs, and topics across Lattice.
V for the SVD model. FVD are tested on the DAVIS2017 validation set. Higher (↑\uparrow) is better; lower (↓\downarrow) is better. Best is bold; second-best is underlined. VBench Appearance (↑\uparrow)
1
0
2
Image-conditioned video diffusion models can now be fine-tuned to produce more realistic motion dynamics and long-term temporal coherence via a novel reward-driven approach that avoids common pitfalls like reward hacking.
