Research map / Detection and segmentation / CNN features
Offset Bin Classification Network for Accurate Object Detection
Heqian Qiu, Hongliang Li, Qingbo Wu, Hengcan Shi
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 |
|---|---|---|---|
| Replace bounding box regression with offset bin classification | Faster R-CNN with ResNet-50-FPN · VOC2007 test | AP: 45.0 -> 49.0 | Improved |
| Replace bounding box regression with offset bin classification | Faster R-CNN with ResNet-50-FPN · MS-COCO test-dev | AP: 36.6 -> 40.5 | Improved |
| Replace regression branches with offset bin classification | Cascade R-CNN with ResNet-101-FPN · VOC2007 test | AP: 51.0 -> 51.9 | Improved |
| Replace regression branches with offset bin classification | Cascade R-CNN with ResNet-101-FPN · MS-COCO test-dev | AP: 42.4 -> 44.4 | Improved |
| Add expectation-based offset prediction (ablation) | Offset bin classification · VOC2007 test | AP: 45.8 -> 47.5 | Improved |
| Add hierarchical focusing offset prediction (ablation) | Offset bin classification · VOC2007 test | AP: 45.8 -> 47.5 | Improved |
| Add both expectation-based and hierarchical focusing offset prediction (ablation) | Offset bin classification · VOC2007 test | AP: 45.8 -> 49.0 | Improved |
| Replace Smooth L1 loss with offset bin classification (ablation) | ResNet-50-FPN detector · VOC2007 test | AP: 45.0 -> 47.5 | Improved |
| Increase Smooth L1 division point from β = 1.0 to β = 1.5 (ablation) | ResNet-50-FPN detector · VOC2007 test | AP: 45.0 -> 44.3 | Hurt |
| Increase Smooth L1 division point from β = 1.0 to β = 2.0 (ablation) | ResNet-50-FPN detector · VOC2007 test | AP: 45.0 -> 44.2 | Hurt |
| Use two hierarchical focusing stages instead of one (ablation) | Offset bin classification · VOC2007 test | AP: 47.5 -> 49.0 | Improved |
| Use three hierarchical focusing stages instead of two (ablation) | Offset bin classification · VOC2007 test | AP: 49.0 -> 48.8 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Cascade R-CNN — Cascade R-CNN: Delving Into High Quality Object Detection (CVPR 2018) · better on 3 of 3 shared results
- Faster R-CNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- OffsetNet: Towards Efficient Multiple Object Tracking, Detection, and Segmentation (TPAMI 2024 · similarity 0.86)
- Learning Globally Optimized Object Detector via Policy Gradient (CVPR 2018 · similarity 0.85)
- Learning Efficient Binarized Object Detectors With Information Compression (TPAMI 2021 · similarity 0.85)
- Learning Rich Features at High-Speed for Single-Shot Object Detection (ICCV 2019 · similarity 0.85)
- Unsupervised Object Detection with Theoretical Guarantees (NeurIPS 2024 · similarity 0.84)
- Improving Single Domain-Generalized Object Detection: A Focus on Diversification and Alignment (CVPR 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.
