Research map / Detection and segmentation / Open-set and few-shot detection
Bending Reality: Distortion-Aware Transformers for Adapting to Panoramic Semantic Segmentation
Check 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 distortion-aware DPE instead of DPT patch embedding (ablation) | Trans4PASS · DensePASS | mIoU: 36.50 -> 45.89 | Helped |
| Use DMLP rather than CycleMLP in the decoder (ablation) | Trans4PASS · DensePASS | mIoU: 40.16 -> 45.89 | Helped |
| Use DMLP rather than ASMLP in the decoder (ablation) | Trans4PASS · DensePASS | mIoU: 42.05 -> 45.89 | Helped |
| Add a DMLP decoder to the PVT-T encoder (ablation) | PVT-T · DensePASS | mIoU: 31.20 -> 35.18 | Helped |
| Add MPA to Trans4PASS-S adaptation (ablation) | Trans4PASS-S · Stanford2D3D-Panoramic fold-1 | mIoU: 48.34 -> 52.15 | Helped |
| Add MPA to Trans4PASS-T adaptation (ablation) | Trans4PASS-T · DensePASS | mIoU: 45.89 -> 51.93 | Helped |
| Add pseudo-label SSL to MPA (ablation) | Trans4PASS-T · DensePASS | mIoU: 51.93 -> 53.26 | Helped |
| Add multi-scale evaluation (ablation) | Trans4PASS-S with MPA + SSL · DensePASS | mIoU: 55.25 -> 56.38 | Helped |
| Set the DPE offset restriction to r=4 rather than no restriction (ablation) | Trans4PASS · DensePASS | mIoU: 45.74 -> 45.89 | Helped |
| Add mutual prototype alignment to source-only adaptation (ablation) | Trans4PASS · Stanford2D3D-Panoramic, 3-fold average | mIoU: 48.1 -> 50.8 | Helped |
Closest papers by meaning
- Semantics Distortion and Style Matter: Towards Source-free UDA for Panoramic Segmentation (CVPR 2024 · similarity 0.91)
- GoodSAM: Bridging Domain and Capacity Gaps via Segment Anything Model for Distortion-aware Panoramic Semantic Segmentation (CVPR 2024 · similarity 0.89)
- Geometric Exploitation for Indoor Panoramic Semantic Segmentation (NeurIPS 2024 · similarity 0.89)
- AdapToPASS: Ambiguity-aware Adaptive Spherical Transformer for Panoramic Semantic Segmentation (NeurIPS 2026 · similarity 0.87)
- PanDA: Towards Panoramic Depth Anything with Unlabeled Panoramas and Mobius Spatial Augmentation (CVPR 2025 · similarity 0.87)
- DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images (NeurIPS 2026 · similarity 0.87)
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.