Research map / Video / Video generation
MCVD - Masked Conditional Video Diffusion for Prediction, Generation, and Interpolation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add past-masking during training (ablation) | MCVD concat · BAIR 1->15 | FVD: 98.8 -> 95.6 | Helped |
| add past-masking during training (ablation) | MCVD spatin · BAIR 2->28 | FVD: 132.1 -> 127.9 | Helped |
| add past-masking during training (ablation) | MCVD concat · BAIR 2->28 | FVD: 120.6 -> 119.0 | Helped |
| add past-masking during training (ablation) | MCVD spatin · BAIR 2->14 | FVD: 94.1 -> 90.5 | Helped |
| add past-future masking (ablation) | MCVD concat · BAIR 1->15 | FVD: 98.8 -> 89.5 | Helped |
| add past-future masking (ablation) | MCVD concat · BAIR 1->15 | PSNR: 18.8 -> 16.9 | Hurt |
| add past-future masking (ablation) | MCVD concat · SMMNIST 5->10 | FVD: 25.63 -> 20.77 | Helped |
| add past-future masking (ablation) | MCVD spatin · SMMNIST 5->10 | FVD: 23.86 -> 36.12 | Hurt |
| add future-only masking (ablation) | MCVD spatin · SMMNIST 5->10 | FVD: 23.86 -> 44.14 | Hurt |
| use SPATIN conditioning instead of concatenation (ablation) | MCVD · Cityscapes 2->28 | FVD: 141.31 -> 184.81 | Hurt |
| use SPATIN conditioning instead of concatenation (ablation) | MCVD · BAIR 2->28 | FVD: 120.6 -> 132.1 | Hurt |
| use SPATIN instead of concat (ablation) | MCVD · SMMNIST 5->10 | FVD: 25.63 -> 23.86 | Mixed |
Closest papers by meaning
- Identifying and Solving Conditional Image Leakage in Image-to-Video Diffusion Model (NeurIPS 2024 · similarity 0.92)
- OSV: One Step is Enough for High-Quality Image to Video Generation (CVPR 2025 · similarity 0.90)
- TRIP: Temporal Residual Learning with Image Noise Prior for Image-to-Video Diffusion Models (CVPR 2024 · similarity 0.89)
- Video Prediction by Modeling Videos as Continuous Multi-Dimensional Processes (CVPR 2024 · similarity 0.89)
- FIFO-Diffusion: Generating Infinite Videos from Text without Training (NeurIPS 2024 · similarity 0.89)
- Efficient Video Diffusion Models via Content-Frame Motion-Latent Decomposition (ICLR 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.