Research map / Video / Video segmentation and tracking
Learning to Generate Time-Lapse Videos Using Multi-Stage Dynamic Generative Adversarial Networks
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 Refine-Net with adversarial ranking loss and Gram-matrix motion features (ablation) | MD-GAN Base-Net · Time-lapse video generation | human preference POS: 70 | Helped |
| Use a 3D U-net generator with encoder-decoder skip connections (ablation) | Base-Net · Time-lapse video generation | visual content detail | Helped |
| Remove skip connections between conv1/deconv6 and conv2/deconv5 (ablation) | Refine-Net · Time-lapse video generation | motion quality | Helped |
| Use adversarial ranking loss to push generated-video motion toward real videos and away from Base-Net output (ablation) | Refine-Net · Time-lapse video generation | motion quality | Helped |
| Use 3D convolutions for multi-frame generation rather than sequential RNN generation | MD-GAN versus RNN-GAN · Human preference comparison | POS: 97 | Helped |
| Use MD-GAN Stage 2 instead of VGAN | Beach · Beach | MSE: 0.0958 -> 0.0422 | Helped |
| Use MD-GAN Stage 2 instead of VGAN | Golf · Golf | SSIM: 0.5133 -> 0.7085 | Helped |
Closest papers by meaning
- End-To-End Time-Lapse Video Synthesis From a Single Outdoor Image (CVPR 2019 · similarity 0.91)
- Generating Videos with Scene Dynamics (NeurIPS 2016 · similarity 0.89)
- Temporal Coherency based Criteria for Predicting Video Frames using Deep Multi-stage Generative Adversarial Networks (NeurIPS 2017 · similarity 0.89)
- Towards Smooth Video Composition (ICLR 2023 · similarity 0.89)
- SnapGen-V: Generating a Five-Second Video within Five Seconds on a Mobile Device (CVPR 2025 · similarity 0.85)
- Autoregressive Adversarial Post-Training for Real-Time Interactive Video Generation (NeurIPS 2025 · similarity 0.85)
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.