Research map / Detection and segmentation / CNN features
Robust Object Detection Under Occlusion With Context-Aware CompositionalNets
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 |
|---|---|---|---|
| Replace RPN proposals with part-based bounding box voting | CA-CompNet, ω=0.5 · OccludedVehiclesDetection | Mean correct AP(%) @IoU0.5: 60.0 -> 76.6 | Helped |
| Add context-aware representation and set context weight to ω=0.2 (ablation) | CA-CompNet via BBV · OccludedVehiclesDetection | Mean correct AP(%) @IoU0.5: 76.6 -> 81.3 | Helped |
| Deactivate context (set ω=0) instead of using ω=0.2 (ablation) | CA-CompNet via BBV · OccludedVehiclesDetection | Mean correct AP(%) @IoU0.5: 81.3 -> 81.4 | No clear effect |
| Set context weight to ω=0.2 instead of ω=0.5 (ablation) | CA-CompNet via BBV · OccludedCOCO, L4 | AP(%) @IoU0.5: 40.4 -> 59.6 | Helped |
| Deactivate context (set ω=0) instead of using ω=0.2 (ablation) | CA-CompNet via BBV · OccludedCOCO, L4 | AP(%) @IoU0.5: 59.6 -> 54.4 | Hurt |
| Train Faster R-CNN with CutOut occlusion augmentation | Faster R-CNN · OccludedVehiclesDetection | Mean correct AP(%) @IoU0.5: 66.5 -> 71.1 | Helped |
| Train Faster R-CNN with occluded images | Faster R-CNN · OccludedCOCO, L4 | AP(%) @IoU0.5: 24.6 -> 45.6 | Helped |
Closest papers by meaning
- Compositional Convolutional Neural Networks: A Deep Architecture With Innate Robustness to Partial Occlusion (CVPR 2020 · similarity 0.93)
- Robust Instance Segmentation Through Reasoning About Multi-Object Occlusion (CVPR 2021 · similarity 0.88)
- DeepVoting: A Robust and Explainable Deep Network for Semantic Part Detection Under Partial Occlusion (CVPR 2018 · similarity 0.88)
- A Bayesian Approach to OOD Robustness in Image Classification (CVPR 2024 · similarity 0.85)
- Multi-Part Object Representations via Graph Structures and Co-Part Discovery (NeurIPS 2026 · similarity 0.85)
- ProxyDet: Synthesizing Proxy Novel Classes via Classwise Mixup for Open-Vocabulary Object Detection (AAAI 2024 · 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.