Research map / Detection and segmentation / CNN features
Exploit All the Layers: Fast and Accurate CNN Object Detector With Scale Dependent Pooling and Cascaded Rejection Classifiers
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 |
|---|---|---|---|
| use scale-dependent pooling from convolutional layers chosen by proposal size | VGG16 Fast R-CNN detector · KITTI train/validation | mAP: 60.3 -> 65.1 | Helped |
| use scale-dependent pooling from convolutional layers chosen by proposal size | VGG16 Fast R-CNN detector · Inner-city | mAP: 62.2 -> 70.7 | Helped |
| use multi-scale inputs with SDP (ablation) | SDP · VOC2007 | mAP: 68.2 -> 69.4 | Helped |
| add cascaded rejection classifiers (ablation) | SDP · KITTI train/validation | mAP: 65.1 -> 64.0 | Hurt |
| add cascaded rejection classifiers (ablation) | SDP · KITTI | proposal evaluation speedup: 3.2× speedup | Helped |
| add cascaded rejection classifiers (ablation) | SDP · Inner-city | mAP: 70.7 -> 69.3 | Hurt |
| add cascaded rejection classifiers (ablation) | SDP · VOC2007 | mAP: 69.4 -> 67.7 | Hurt |
| fine-tune the network using CRC-surviving proposals (ablation) | SDP+CRC · Inner-city | mAP: 69.3 -> 70.4 | Helped |
| combine CRC with truncated SVD (ablation) | SDP · KITTI | total speedup: 2.37 | Helped |
Closest papers by meaning
- Cascade R-CNN: Delving Into High Quality Object Detection (CVPR 2018 · similarity 0.87)
- MaxpoolNMS: Getting Rid of NMS Bottlenecks in Two-Stage Object Detectors (CVPR 2019 · similarity 0.87)
- Feature Pyramid Networks for Object Detection (CVPR 2017 · similarity 0.87)
- Weakly Supervised Cascaded Convolutional Networks (CVPR 2017 · similarity 0.86)
- CRAFT Objects From Images (CVPR 2016 · similarity 0.86)
- Scale-Equalizing Pyramid Convolution for Object Detection (CVPR 2020 · 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.