Research map / Detection and segmentation / CNN features
Spherical Criteria for Fast and Accurate 360° Object Detection
Read the paperCheck 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 |
|---|---|---|---|
| use spherical criteria (SphBB/SphIoU) instead of ERPBB/ERPIoU (ablation) | Rep R-CNN · VOC360 | mAP: 30.61 -> 71.65 | Helped |
| use spherical criteria instead of ERPBB/ERPIoU (ablation) | Rep R-CNN · COCO-Men | mAP: 37.79 -> 81.48 | Helped |
| use spherical criteria instead of circular BB/CirIoU (ablation) | Rep R-CNN · VOC360 | mAP: 24.48 -> 71.65 | Helped |
| use spherical criteria instead of TanBB/PolyIoU (ablation) | Rep R-CNN · VOC360 | mAP: 58.35 -> 71.65 | Helped |
| use spherical criteria instead of TanBB/PolyIoU (ablation) | Rep R-CNN · COCO-Men | mAP: 65.17 -> 81.48 | Helped |
| use spherical criteria instead of unbiased UnbIoU (ablation) | Rep R-CNN · COCO-Men | mAP: 83.89 -> 81.48 | Hurt |
| use spherical criteria instead of unbiased UnbIoU (ablation) | Rep R-CNN · VOC360 | mAP: 72.47 -> 71.65 | Hurt |
| increase number of PSPs n from 10 to 50 (ablation) | Rep R-CNN · COCO-Men | mAP: 65.43 -> 81.48 | Helped |
| increase number of PSPs n from 20 to 100 (ablation) | Rep R-CNN · VOC360 | mAP: 71.88 -> 71.57 | No clear effect |
| share backbone weights between SphRPN and RepNet (ablation) | Rep R-CNN · VOC360 | mAP | Hurt |
| Rep R-CNN vs strongest baseline Multi-projection YOLO | Rep R-CNN · VOC360 | mAP: 54.29 -> 71.88 | Helped |
Closest papers by meaning
- Unbiased IoU for Spherical Image Object Detection (AAAI 2022 · similarity 0.90)
- Gaussian Label Distribution Learning for Spherical Image Object Detection (CVPR 2023 · similarity 0.87)
- SpherePHD: Applying CNNs on a Spherical PolyHeDron Representation of 360deg Images (CVPR 2019 · similarity 0.87)
- Unified Spherical Frontend: Learning Rotation-Equivariant Representations of Spherical Images from Any Camera (CVPR 2026 · similarity 0.87)
- Cubify Anything: Scaling Indoor 3D Object Detection (CVPR 2025 · similarity 0.85)
- UNION: Unsupervised 3D Object Detection using Object Appearance-based Pseudo-Classes (NeurIPS 2024 · similarity 0.84)
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.