Research map / Detection and segmentation / CNN features
A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection
Zhaowei Cai, Quanfu Fan, Rogerio S. Feris, Nuno Vasconcelos
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 |
|---|---|---|---|
| add input upsampling (height 384 to 576) (ablation) | MS-CNN · KITTI val Cars Moderate | AP: 80.63 -> 88.14 | Improved |
| add input upsampling (height 384 to 576) (ablation) | MS-CNN · KITTI val Pedestrians Moderate | AP: 68.37 -> 70.77 | Improved |
| raise input height beyond 2x with deconvolution (576 to 768) (ablation) | MS-CNN with deconvolution · KITTI val Cars Moderate | AP: 89.12 -> 88.83 | No clear effect |
| replace random sampling with bootstrapping (ablation) | MS-CNN h576 · KITTI val Pedestrians Moderate | AP: 65.91 -> 70.77 | Improved |
| replace mixture sampling with bootstrapping (ablation) | MS-CNN h576 · KITTI val Pedestrians Moderate | AP: 70.49 -> 70.77 | No clear effect |
| add deconvolution feature upsampling (ablation) | MS-CNN h384 · KITTI val Cars Moderate | AP: 80.63 -> 87.93 | Improved |
| add deconvolution feature upsampling (ablation) | MS-CNN h576 · KITTI val Pedestrians Moderate | AP: 70.77 -> 72.49 | Improved |
| add context encoding (ablation) | MS-CNN h576 · KITTI val Pedestrians Moderate | AP: 70.77 -> 71.45 | Improved |
| add dimensionality reduction conv to context model (ablation) | MS-CNN h576-ctx · KITTI val Pedestrians Moderate | AP: 71.45 -> 72.13 | No clear effect |
| combine detectors at multiple layers instead of single det-8 layer (ablation) | MS-CNN proposal network · KITTI val car recall (all scales) | recall: 0.6486 -> 0.9611 | Improved |
| add detection sub-network multi-task training to proposal learning (ablation) | MS-CNN · KITTI val proposal recall | recall | Improved |
| MS-CNN versus 3DOP | MS-CNN · KITTI test Pedestrians Moderate | AP: 67.47 -> 73.70 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Faster-RCNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 3 of 3 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
- Feature Pyramid Networks for Object Detection (CVPR 2017 · similarity 0.91)
- SCNN: A General Distribution Based Statistical Convolutional Neural Network with Application to Video Object Detection (AAAI 2019 · similarity 0.89)
- SAN: Learning Relationship between Convolutional Features for Multi-Scale Object Detection (ECCV 2018 · similarity 0.89)
- R-FCN: Object Detection via Region-based Fully Convolutional Networks (NeurIPS 2016 · similarity 0.89)
- Deep Learning for Generic Object Detection: A Survey (IJCV 2019 · similarity 0.89)
- Region-Based Convolutional Networks for Accurate Object Detection and Segmentation (TPAMI 2016 · similarity 0.89)
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.
