Research map / Video / Video segmentation and tracking
SimVP: Simpler Yet Better Video Prediction
Check 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 |
|---|---|---|---|
| Use a simple end-to-end CNN video predictor with MSE loss | video prediction · Moving MNIST | MSE: 24.4 -> 23.8 | Helped |
| Use a simple end-to-end CNN video predictor with MSE loss | video prediction · Human3.6 | MSE / 10: 36.9 -> 31.6 | Helped |
| Use a simple end-to-end CNN video predictor with MSE loss | video prediction · TrafficBJ | MAE: 16.2 -> 16.2 | No clear effect |
| Use a simple end-to-end CNN video predictor | video prediction · Caltech Pedestrian (10 → 1) | SSIM: 0.927 -> 0.940 | Helped |
| Use a simple end-to-end CNN video predictor | video prediction · KTH (10 → 40) | PSNR: 27.56 -> 32.93 | Helped |
| Add spatial UNet shortcut (ablation) | SimVP · Moving MNIST | MSE: 41.7 -> 41.7 | Helped |
| Add temporal UNet shortcut (ablation) | SimVP · Moving MNIST | MSE: 41.5 -> 41.7 | No clear effect |
| Use grouped convolution (4 groups rather than 1) (ablation) | SimVP · Moving MNIST | MSE: 44.8 -> 41.7 | Helped |
| Use group normalization rather than batch normalization (ablation) | SimVP · Moving MNIST | MSE: 41.0 -> 41.7 | No clear effect |
| Use multi-scale kernels rather than a single kernel of size 3 (ablation) | SimVP translator · Moving MNIST | MSE: 58.9 -> 41.7 | Helped |
| Use multi-scale kernels rather than a single kernel of size 5 (ablation) | SimVP translator · Moving MNIST | MSE: 51.1 -> 41.7 | Helped |
| Use multi-scale kernels rather than a single kernel of size 7 (ablation) | SimVP translator · Moving MNIST | MSE: 49.1 -> 41.7 | Helped |
Closest papers by meaning
- Stochastic Variational Video Prediction (ICLR 2018 · similarity 0.86)
- SyncVP: Joint Diffusion for Synchronous Multi-Modal Video Prediction (CVPR 2025 · similarity 0.85)
- Stochastic Backpropagation: A Memory Efficient Strategy for Training Video Models (CVPR 2022 · similarity 0.85)
- VideoTitans: Scalable Video Prediction with Integrated Short- and Long-term Memory (NeurIPS 2025 · similarity 0.84)
- Neural-Centric Video Processing Pipeline for Unified Multi-Task Inference (CVPR 2026 · similarity 0.84)
- Sparse Video-Gen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsity (ICML 2025 · similarity 0.83)
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.