Research map / Detection and segmentation / Open-set and few-shot detection
Capturing Omni-Range Context for Omnidirectional Segmentation
Kailun Yang, Jiaming Zhang, Simon Reiss, Xinxin Hu, Rainer Stiefelhagen
Check 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 efficient concurrent attention to the omni-supervised ERF-PSPNet (ablation) | ERF-PSPNet · WildPASS | mIoU: 66.8% -> 67.7% | Improved |
| Add multi-space prediction fusion to ECANet (ablation) | ECANet · WildPASS | mIoU: 67.7% -> 69.0% | Improved |
| Train with multi-source data instead of Vistas only (ablation) | ERF-PSPNet · WildPASS | mIoU: 34.0% -> 39.8% | Improved |
| Use multi-source rather than single-source fine-tuning (ablation) | ERF-PSPNet · WildPASS | mIoU: 41.1% -> 60.7% | Improved |
| Distill from worldwide panoramas rather than Pittsburgh panoramas (ablation) | Omni-supervised ERF-PSPNet · WildPASS | mIoU: 56.1% -> 66.8% | Improved |
| Use HSA and PSA together instead of HSA alone (ablation) | ECANet · WildPASS | mIoU: 39.4% -> 39.7% | Improved |
| Add HSA to PSA (ablation) | ECANet · WildPASS | mIoU: 67.2% -> 67.7% | Improved |
| Use 4 HSA segments rather than 2 (ablation) | ECANet · WildPASS | mIoU: 66.8% -> 67.4% | Improved |
| Use 8 HSA segments rather than 4 (ablation) | ECANet · WildPASS | mIoU: 67.4% -> 66.7% | Hurt |
| Use calibrated top-two confidence for fusion rather than maximum probability (ablation) | ECANet · WildPASS | mIoU: 68.8% -> 69.0% | Improved |
| Use calibrated top-two confidence for fusion rather than minimum variance (ablation) | ECANet · WildPASS | mIoU: 68.4% -> 69.0% | Improved |
Closest papers by meaning
- A Survey of Representation Learning, Optimization Strategies, and Applications for Omnidirectional Vision (IJCV 2025 · similarity 0.89)
- FoveaNet: Perspective-Aware Urban Scene Parsing (ICCV 2017 · similarity 0.88)
- Unifying Training and Inference for Panoptic Segmentation (CVPR 2020 · similarity 0.88)
- OmniHD-Scenes: A Next-Generation Multimodal Dataset for Autonomous Driving (TPAMI 2026 · similarity 0.87)
- OmniSegmentor: A Flexible Multi-Modal Learning Framework for Semantic Segmentation (NeurIPS 2025 · similarity 0.86)
- OA-CNNs: Omni-Adaptive Sparse CNNs for 3D Semantic Segmentation (CVPR 2024 · similarity 0.86)
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.
