Research map / Time series / Foundation models for time series
Panoptic Segmentation of Satellite Image Time Series With Convolutional Temporal Attention Networks
Vivien Sainte Fare Garnot, Loic Landrieu
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 |
|---|---|---|---|
| add interpolated multi-level temporal attention masks instead of temporal average of skip connections (ablation) | U-TAE · PASTIS semantic segmentation (5-fold CV) | mIoU: 63.1 vs 58.3 | Improved |
| add 1x1 conv after temporal average -> still worse than attention masks (ablation) | U-TAE · PASTIS semantic segmentation (5-fold CV) | mIoU: 63.1 vs 58.9 | Improved |
| add channel grouping of attention masks (G heads) instead of one averaged mask (ablation) | U-TAE · PASTIS semantic segmentation (5-fold CV) | mIoU: 63.1 vs 60.1 | Improved |
| use group normalization instead of batch normalization in encoder (ablation) | U-TAE · PASTIS semantic segmentation (5-fold CV) | mIoU: 63.1 vs 36.0 | Improved |
| use full time series instead of single August date (ablation) | U-TAE · PASTIS semantic segmentation (5-fold CV) | mIoU: 63.1 vs 28.3 | Improved |
| use full time series instead of single May date (ablation) | U-TAE · PASTIS semantic segmentation (5-fold CV) | mIoU: 63.1 vs 20.6 | Improved |
| replace U-BiConvLSTM-style recurrent encoder with U-TAE in PaPs | PaPs · PASTIS panoptic segmentation | PQ: 35.6 -> 43.8 | Improved |
| vary shape patch size S from 16 to 24 (ablation) | PaPs · PASTIS panoptic segmentation | PQ: 43.8 -> 41.3 | No clear effect |
| vary shape patch size S from 16 to 8 (ablation) | PaPs · PASTIS panoptic segmentation | PQ: 43.8 -> 42.7 | No clear effect |
| add residual CNN shape prediction instead of element-wise multiplication (ablation) | PaPs · PASTIS panoptic segmentation | SQ: 74.6 -> 81.5 | Improved |
| use time series instead of single image (ablation) | PaPs · PASTIS panoptic segmentation | PQ: 14.1 -> 43.8 | Improved |
| U-TAE vs strongest baseline 3D-Unet | semantic seg · PASTIS semantic segmentation (5-fold CV) | mIoU: 58.4 -> 63.1 | Improved |
Closest papers by meaning
- Exact: Exploring Space-Time Perceptive Clues for Weakly Supervised Satellite Image Time Series Semantic Segmentation (CVPR 2025 · similarity 0.91)
- Satellite Image Time Series Classification With Pixel-Set Encoders and Temporal Self-Attention (CVPR 2020 · similarity 0.90)
- SatSynth: Augmenting Image-Mask Pairs through Diffusion Models for Aerial Semantic Segmentation (CVPR 2024 · similarity 0.86)
- Efficient Representation Learning of Satellite Image Time Series and Their Fusion for Spatiotemporal Applications (AAAI 2024 · similarity 0.86)
- TEOChat: A Large Vision-Language Assistant for Temporal Earth Observation Data (ICLR 2025 · similarity 0.85)
- Towards Open-Vocabulary Remote Sensing Image Semantic Segmentation (AAAI 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.
