Research map / Detection and segmentation / CNN features
Scale-Transferrable Object Detection
Peng Zhou, Bingbing Ni, Cong Geng, Jianguo Hu, Yi Xu
Read the paperCheck 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 |
|---|---|---|---|
| Use STDN300 instead of SSD300 | Object detection on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 77.5 -> 78.1 | Improved |
| Use STDN513 instead of SSD512 | Object detection on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 79.5 -> 80.9 | Improved |
| Use STDN321 instead of DSSD321 | Object detection on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 78.6 -> 79.3 | Improved |
| Use STDN513 instead of DSSD513 | Object detection on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 81.5 -> 80.9 | Hurt |
| Use STDN300 instead of SSD300 | Object detection on MS COCO · MS COCO test-dev2015 | AP (IoU 0.5:0.95): 25.1 -> 28.0 | Improved |
| Use STDN513 instead of SSD512 | Object detection on MS COCO · MS COCO test-dev2015 | AP (IoU 0.5:0.95): 28.8 -> 31.8 | Improved |
| Use STDN321 instead of DSSD321 | Object detection on MS COCO · MS COCO test-dev2015 | AP (IoU 0.5:0.95): 28.0 -> 28.0 | No clear effect |
| Use STDN513 instead of DSSD513 | Object detection on MS COCO · MS COCO test-dev2015 | AP (IoU 0.5:0.95): 33.2 -> 31.8 | Hurt |
| Add STM to DenseNet-169 using SSD-style feature maps (ablation) | DenseNet-169 detector on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 73.8 -> 76.6 | Improved |
| Add the stem block (ablation) | DenseNet-169 + STM on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 76.6 -> 78.1 | Improved |
| Replace DenseNet-169 with Inception-v3 (ablation) | STM detector on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 78.1 -> 74.9 | Hurt |
Closest papers by meaning
- Scaling Object Detection by Transferring Classification Weights (ICCV 2019 · similarity 0.88)
- POD: Practical Object Detection With Scale-Sensitive Network (ICCV 2019 · similarity 0.87)
- SAN: Learning Relationship between Convolutional Features for Multi-Scale Object Detection (ECCV 2018 · similarity 0.87)
- A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection (ECCV 2016 · similarity 0.87)
- An Analysis of Scale Invariance in Object Detection SNIP (CVPR 2018 · similarity 0.86)
- ScaleNet: A Shallow Architecture for Scale Estimation (CVPR 2022 · 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.
