Research map / Video / Video segmentation and tracking
Learning-based Video Motion Magnification
Tae-Hyun Oh, Ronnachai Jaroensri, Changil Kim, Mohamed Elgharib, Fr'edo Durand, William T. Freeman, Wojciech Matusik
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 a non-linear manipulator instead of a linear one (ablation) | learned motion magnification network · real videos (Fig. 3) | edge blur and noise | Improved |
| apply a temporal filter to the learned shape representation | learned motion magnification network · drum sequence (Fig. 5) | blurring artifacts | Improved |
| replace perceptual or adversarial loss for l1 loss (ablation) | learned motion magnification network · training | output quality | No clear effect |
| add static scenes to the synthetic training data (ablation) | learned motion magnification network · training data | noise vs motion discrimination | Improved |
| add blurred-background and background-only examples (ablation) | learned motion magnification network · low contrast and large objects | generalization | Improved |
| use learned representation instead of complex steerable pyramid | phase-based motion magnification · baby and balance sequences | ringing artifacts | Improved |
| use learned representation instead of phase-based under noise | phase-based motion magnification · synthetic noise test (Fig. 8) | quality vs noise factor | Improved |
| raise magnification factor above 20x with temporal filter | learned motion magnification network · synthetic videos (Fig. 11) | blur and color artifacts | Hurt |
Closest papers by meaning
- Self-Supervised Motion Magnification by Backpropagating Through Optical Flow (NeurIPS 2023 · similarity 0.90)
- Event-Based Motion Magnification (ECCV 2024 · similarity 0.90)
- Learning-based Axial Video Motion Magnification (ECCV 2024 · similarity 0.90)
- Multi Domain Learning for Motion Magnification (CVPR 2023 · similarity 0.90)
- Frequency Decoupling for Motion Magnification via Multi-Level Isomorphic Architecture (CVPR 2024 · similarity 0.89)
- EulerMormer: Robust Eulerian Motion Magnification via Dynamic Filtering within Transformer (AAAI 2024 · similarity 0.88)
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.
