Research map / Detection and segmentation / Weakly supervised detection and segmentation
Instance Mining with Class Feature Banks for Weakly Supervised Object Detection
Read the paperCheck 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 |
|---|---|---|---|
| Add FGIM to MIL (ablation) | WSOD detector · PASCAL VOC 2007 | mAP (%): 34.8 -> 38.0 | Helped |
| Add OIR to MIL (ablation) | WSOD detector · PASCAL VOC 2007 | mAP (%): 34.8 -> 52.1 | Helped |
| Add FGIM to MIL + OIR (ablation) | WSOD detector · PASCAL VOC 2007 | mAP (%): 52.1 -> 54.3 | Helped |
| Use FIFO instead of weighting for CFB updates (ablation) | IM-CFB · PASCAL VOC 2007 | mAP (%): 54.3 -> 52.6 | Hurt |
| Use CFB weighting updates rather than no CFB (ablation) | MIL + OIR · PASCAL VOC 2007 | mAP (%): 52.1 -> 54.3 | Helped |
| Set CFB sub-bank length K to 2 (ablation) | IM-CFB · PASCAL VOC 2007 | mAP (%): 54.3 -> 51.5 | Hurt |
| Set CFB sub-bank length K to 4 (ablation) | IM-CFB · PASCAL VOC 2007 | mAP (%): 54.3 -> 52.7 | Hurt |
| Set CFB sub-bank length K to 8 (ablation) | IM-CFB · PASCAL VOC 2007 | mAP (%): 54.3 -> 53.1 | Hurt |
| Set CFB sub-bank length K to 10 (ablation) | IM-CFB · PASCAL VOC 2007 | mAP (%): 54.3 -> 52.3 | Hurt |
| Set CFB sub-bank length K to 20 (ablation) | IM-CFB · PASCAL VOC 2007 | mAP (%): 54.3 -> 51.3 | Hurt |
| Use IM-CFB rather than SLV | WSOD detector · PASCAL VOC 2007 test | mAP (%): 53.5 -> 54.3 | Helped |
| Use IM-CFB rather than SLV | WSOD detector · PASCAL VOC 2007 trainval | CorLoc (%): 71.0 -> 70.7 | Hurt |
Closest papers by meaning
- Object Instance Mining for Weakly Supervised Object Detection (AAAI 2020 · similarity 0.93)
- UWSOD: Toward Fully-Supervised-Level Capacity Weakly Supervised Object Detection (NeurIPS 2020 · similarity 0.90)
- C-MIL: Continuation Multiple Instance Learning for Weakly Supervised Object Detection (CVPR 2019 · similarity 0.90)
- Instance-Aware, Context-Focused, and Memory-Efficient Weakly Supervised Object Detection (CVPR 2020 · similarity 0.89)
- Weakly Supervised Open-Vocabulary Object Detection (AAAI 2024 · similarity 0.89)
- Weakly Supervised Few-Shot Object Detection with DETR (AAAI 2024 · similarity 0.88)
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.