Research map / Video / Video segmentation and tracking
A New Large Scale Dynamic Texture Dataset with Application to ConvNet Understanding
Isma Hadji, Richard P. Wildes
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 MSOE-filtered input instead of optical flow for the motion stream (ablation) | Flow stream · DTDB-Dynamics | accuracy: 72.6 -> 80.1 | Improved |
| use MSOE-filtered input instead of optical flow for the motion stream (ablation) | Flow stream · DTDB-Appearance | accuracy: 64.8 -> 72.2 | Improved |
| fuse RGB appearance stream with the MSOE stream (ablation) | MSOE stream · DTDB-Dynamics | accuracy: 80.1 -> 84.0 | Improved |
| fuse RGB appearance stream with the MSOE stream (ablation) | MSOE stream · DTDB-Appearance | accuracy: 72.2 -> 80.0 | Improved |
| pretrain on DTDB-dynamics instead of UCF-101 | MSOE-two-stream · YUVL-1 feature transfer | accuracy: 80.8 -> 90.7 | Improved |
| pretrain on DTDB-appearance instead of UCF-101 | MSOE-two-stream · YUP++ feature transfer | accuracy: 74.6 -> 89.6 | Improved |
| pretrain on DTDB-appearance instead of DTDB-dynamics | MSOE-two-stream · YUVL-1 feature transfer | accuracy: 90.7 -> 89.8 | Hurt |
| pretrain on DTDB-appearance instead of DTDB-dynamics | MSOE-two-stream · YUP++ feature transfer | accuracy: 87.7 -> 89.6 | Improved |
| finetune pretrained MSOE-two-stream on DTDB-appearance | MSOE-two-stream · YUP++ entire | accuracy: 89.6 -> 91.8 | Improved |
| finetune pretrained C3D on DTDB-dynamics instead of using DTDB-trained-from-scratch features | C3D · YUVL-1 | accuracy: 83.3 -> 89.1 | Improved |
| camera motion present (moving vs static camera) | MSOE-two-stream trained on DTDB-appearance · YUP++ feature transfer | accuracy: 94.7 -> 83.2 | Hurt |
| use hand-crafted SOE-Net instead of learned MSOE-two-stream | MSOE-two-stream · DTDB-Dynamics | accuracy: 84.0 -> 86.8 | Improved |
Closest papers by meaning
- Two-Stream Convolutional Networks for Dynamic Texture Synthesis (CVPR 2018 · similarity 0.90)
- DiVa-360: The Dynamic Visual Dataset for Immersive Neural Fields (CVPR 2024 · similarity 0.85)
- Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting (ICLR 2024 · similarity 0.84)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.83)
- DyST: Towards Dynamic Neural Scene Representations on Real-World Videos (ICLR 2024 · similarity 0.83)
- OSN: Infinite Representations of Dynamic 3D Scenes from Monocular Videos (ICML 2024 · 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.
