Research map / Video / Video segmentation and tracking
Folded Recurrent Neural Networks for Future Video Prediction
Marc Oliu, Javier Selva, Sergio Escalera
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 state sharing between encoder and decoder instead of bridge connections (ablation) | fRNN · MMNIST | PSNR: 13.857 -> 21.386 | Improved |
| use state sharing instead of bridge connections (ablation) | fRNN · KTH | PSNR: 31.268 -> 29.299 | Hurt |
| use state sharing instead of bridge connections (ablation) | fRNN · UCF101 | PSNR: 23.558 -> 23.872 | Improved |
| use state sharing instead of bridge connections (ablation) | fRNN · KTH | DSSIM: 0.05945 -> 0.07251 | Hurt |
| pre-train RLadder on KTH to learn an identity mapping before MMNIST (ablation) | RLadder · MMNIST | PSNR / MSE / DSSIM | Improved |
| predict without feedback of previous predictions (fRNN) instead of feeding predictions back | fRNN vs Villegas · MMNIST | PSNR: 13.857 -> 21.386 | Improved |
| predict without feedback of previous predictions (fRNN) instead of feeding predictions back | fRNN vs Villegas · KTH | PSNR: 30.946 -> 29.299 | Hurt |
| predict without feedback of previous predictions (fRNN) instead of feeding predictions back | fRNN vs Villegas · UCF101 | PSNR: 23.457 -> 23.872 | Improved |
| remove the last 2 bGRU layers of trained fRNN (ablation) | fRNN · MMNIST (qualitative) | visual prediction quality | No clear effect |
| remove the third deepest bGRU layer (ablation) | fRNN · MMNIST (qualitative) | visual prediction quality | Hurt |
| remove all but a single bGRU layer (ablation) | fRNN · MMNIST (qualitative) | visual prediction quality | Hurt |
Closest papers by meaning
- Learning to Generate Long-term Future via Hierarchical Prediction (ICML 2017 · similarity 0.88)
- Recurrent Neural Networks With Intra-Frame Iterations for Video Deblurring (CVPR 2019 · similarity 0.87)
- Recurrent Autoregressive Diffusion: Global Memory Meets Local Attention (ECCV 2026 · similarity 0.87)
- High Fidelity Video Prediction with Large Stochastic Recurrent Neural Networks (NeurIPS 2019 · similarity 0.86)
- PredRNN++: Towards A Resolution of the Deep-in-Time Dilemma in Spatiotemporal Predictive Learning (ICML 2018 · similarity 0.86)
- Flexible Spatio-Temporal Networks for Video Prediction (CVPR 2017 · similarity 0.86)
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.
