Research map / Detection and segmentation / Open-set and few-shot detection
Towards Improving Calibration in Object Detection Under Domain Shift
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 TCD auxiliary loss to FCOS | FCOS · CS to CS-foggy OOD | D-ECE: 18.4 -> 5.5 | Helped |
| add TCD auxiliary loss to FCOS | FCOS · Sim10K to CS | AP@0.5: 37.6 -> 42.4 | Helped |
| apply post-hoc temperature scaling to FCOS | FCOS · in-domain and OOD detections | D-ECE: 15.1 -> 25.3 | Hurt |
| apply post-hoc temperature scaling to Faster-RCNN | Faster-RCNN · Sim10K in-domain | D-ECE: 23.7 -> 15.0 | Helped |
| add TCD to Faster-RCNN | Faster-RCNN · CS to BDD100K | D-ECE: 16.3 -> 9.6 | Helped |
| add TCD to EPM domain-adaptive detector | EPM · Sim10K to CS | D-ECE: 16.0 -> 9.9 | Helped |
| add TCD to SWDA two-stage adaptive detector | SWDA · Sim10K to CS | AP@0.5: 40.9 -> 40.0 | Hurt |
| add TCD to SSAL(UGPL) | SSAL(UGPL) · Sim10K to CS | D-ECE: 13.6 -> 8.5 | Helped |
| add implicit calibration technique (ICT) to SSAL(UGPL) (ablation) | SSAL(UGPL) · Sim10K to CS | D-ECE: 13.6 -> 12.7 | Helped |
| add ICT on top of TCD (ablation) | SSAL(UGPL)+TCD · CS to CS-foggy | D-ECE: 19.1 -> 16.7 | Helped |
| add detection component d_det to TCD (ablation) | TCD on FCOS · CS to CS-foggy OOD | D-ECE: 8.1 -> 5.5 | Helped |
| add classification component d_cls to TCD (ablation) | TCD on FCOS · CS to CS-foggy OOD | D-ECE: 13.3 -> 5.5 | Helped |
Closest papers by meaning
- Bridging Precision and Confidence: A Train-Time Loss for Calibrating Object Detection (CVPR 2023 · similarity 0.93)
- Improving Single Domain-Generalized Object Detection: A Focus on Diversification and Alignment (CVPR 2024 · similarity 0.87)
- Towards Unsupervised Model Selection for Domain Adaptive Object Detection (NeurIPS 2024 · similarity 0.87)
- What How and When Should Object Detectors Update in Continually Changing Test Domains? (CVPR 2024 · similarity 0.86)
- Decoupling Feature Extraction and Classification Layers for Calibrated Neural Networks (ICML 2024 · similarity 0.86)
- Benchmarking Object Detectors under Real-World Distribution Shifts in Satellite Imagery (CVPR 2025 · 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.