Research map / Detection and segmentation / CNN features
BiDet: An Efficient Binarized Object Detector
Ziwei Wang, Ziyi Wu, Jiwen Lu, Jie Zhou
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 |
|---|---|---|---|
| Apply the information bottleneck and sparse object priors instead of direct binarization | SSD300 with VGG16 · PASCAL VOC 2007 test | mAP: 50.2 -> 52.4; +2.2% | Improved |
| Apply the information bottleneck and sparse object priors instead of direct binarization | Faster R-CNN with ResNet-18 · PASCAL VOC 2007 test | mAP: 48.4 -> 50.0; +1.6% | Improved |
| Add extra shortcuts to BiDet | SSD300 with VGG16 · PASCAL VOC 2007 test | mAP: 63.8 -> 66.0; 2.2% | Improved |
| Add extra shortcuts to BiDet | Faster R-CNN with ResNet-18 · PASCAL VOC 2007 test | mAP: 58.2 -> 59.5; 1.3% | Improved |
| Apply the information bottleneck and sparse object priors instead of direct binarization | SSD300 with VGG16 · COCO 2014 minival | mAP@[.5, .95]: 8.1 -> 9.8; 1.7% | Improved |
| Apply the information bottleneck and sparse object priors instead of direct binarization | Faster R-CNN with ResNet-18 · COCO 2014 minival | mAP@[.5, .95]: 10.4 -> 12.1; 1.7% | Improved |
| Add extra shortcuts to BiDet | SSD300 with VGG16 · COCO 2014 minival | mAP@[.5, .95]: 11.2 -> 13.2; 2.0% | Improved |
| Add extra shortcuts to BiDet | Faster R-CNN with ResNet-18 · COCO 2014 minival | mAP@[.5, .95]: 14.4 -> 15.7; 1.3% | Improved |
| Use a medium information-bottleneck weight β rather than a small or large one (ablation) | BiDet with SSD300 and VGG16 · PASCAL VOC | mAP | Improved |
| Use a medium information-bottleneck weight β rather than a small or large one (ablation) | BiDet with SSD300 and VGG16 · PASCAL VOC | false positives | Improved |
| Use a medium sparse-prior weight γ rather than a small or large one (ablation) | BiDet with SSD300 and VGG16 · PASCAL VOC | mAP | Improved |
| Increase γ excessively (ablation) | BiDet with SSD300 and VGG16 · PASCAL VOC | false negatives | Hurt |
Later papers that compared with this method
1 later accepted paper reports BiDet as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Recurrent Bilinear Optimization for Binary Neural Networks (ECCV 2022 · better on 4 of 4 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
- Learning Efficient Binarized Object Detectors With Information Compression (TPAMI 2021 · similarity 0.97)
- EfficientDet: Scalable and Efficient Object Detection (CVPR 2020 · similarity 0.86)
- BiViT: Extremely Compressed Binary Vision Transformers (ICCV 2023 · similarity 0.86)
- Sparse R-CNN: An End-to-End Framework for Object Detection (TPAMI 2023 · similarity 0.86)
- Deep Learning for Generic Object Detection: A Survey (IJCV 2019 · similarity 0.85)
- Cascade R-CNN: Delving Into High Quality Object Detection (CVPR 2018 · 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.
