Research map / Video / Video generation
MoCoGAN: Decomposing Motion and Content for Video Generation
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 |
|---|---|---|---|
| decompose latent codes into fixed content and recurrently generated motion | MoCoGAN · Facial Expressions | ACD: 0.322 -> 0.201 (VGAN); 0.305 -> 0.201 (TGAN) | Helped |
| decompose latent codes into fixed content and recurrently generated motion | MoCoGAN · Shape Motion | ACD: 5.02 -> 1.79 (VGAN); 2.08 -> 1.79 (TGAN) | Helped |
| use MoCoGAN instead of competing video generators | MoCoGAN · UCF101 | Inception score: 8.18 ± .05 -> 12.42 ± .03 (VGAN); 11.85 ± .07 -> 12.42 ± .03 (TGAN) | Helped |
| condition the recurrent motion network on action category rather than the image generator (ablation) | MoCoGAN categorical facial expression generation · Categorical facial expression generation | MCS: 0.472 -> 0.491 (with DI); 0.355 -> 0.581 (without DI) | Mixed |
| include the image discriminator DI (ablation) | MoCoGAN categorical facial expression generation · Categorical facial expression generation | MCS: 0.355 -> 0.472 (zA → GI); 0.581 -> 0.491 (zA → RM) | Mixed |
| increase the content dimension while reducing the motion dimension (ablation) | MoCoGAN facial expression generation · Facial expression generation | MCS: 0.581, 0.525, 0.425, 0.141, 0.245 | Mixed |
| use MoCoGAN instead of C-VGAN | MoCoGAN image-to-video translation · Tai-Chi | User preference, %: 66.9 / 33.1 | Helped |
| use MoCoGAN instead of MCNET | MoCoGAN image-to-video translation · Tai-Chi | User preference, %: 65.6 / 34.4 | Helped |
Closest papers by meaning
- Enhancing Motion in Text-to-Video Generation with Decomposed Encoding and Conditioning (NeurIPS 2024 · similarity 0.85)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.85)
- Decouple Content and Motion for Conditional Image-to-Video Generation (AAAI 2024 · similarity 0.85)
- LAMP: Learn A Motion Pattern for Few-Shot Video Generation (CVPR 2024 · similarity 0.85)
- MotionDreamer: One-to-Many Motion Synthesis with Localized Generative Masked Transformer (ICLR 2025 · similarity 0.84)
- OSV: One Step is Enough for High-Quality Image to Video Generation (CVPR 2025 · similarity 0.84)
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.