AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Simple Multi-Dataset Detection

Xingyi Zhou, Vladlen Koltun, Philipp Krähenbühl

CVPR 2022 · Detection and segmentation › Open-set and few-shot detection · approach: Open-set and unknown object detection · 90 citations (OpenAlex)

Check 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 changedWhereMeasuredEffect
Use dataset-specific heads and objectives with a shared backbone instead of simple dataset concatenationMulti-dataset CascadeRCNN · Mean of COCO, Objects365, OpenImagesmAP: 33.2 -> 41.7Improved
Sample datasets uniformly rather than in proportion to dataset size (ablation)Multi-dataset detector · Mean of COCO, Objects365, OpenImagesmAP: 33.2 -> 34.5Improved
Use class-aware sampling within Objects365 and OpenImages (ablation)Multi-dataset detector · Objects365mAP: 16.5 -> 18.5Improved
Combine uniform dataset sampling with class-aware sampling (ablation)Multi-dataset detector · Mean of COCO, Objects365, OpenImagesmAP: 38.5 -> 40.6Improved
Add a hierarchy-aware sigmoid cross-entropy loss for OpenImages (ablation)Partitioned detector · OpenImagesmAP: +2.7 mAPImproved
Extend the partitioned detector training schedule from 2× to 6× (ablation)Partitioned detector · COCOmAP: 41.8 -> 44.6Improved
Use an AP-based merge objective rather than GloVe word distance (ablation)Unified label space · Mean of COCO, Objects365, OpenImagesmAP: 41.4{ $\pm$0.05} -> 41.9{ $\pm$0.02}Improved
Learn the taxonomy from detector behavior rather than using the expert taxonomy (ablation)Unified detector · Mean of COCO, Objects365, OpenImagesmAP: 41.6{ $\pm$0.04} -> 41.9{ $\pm$0.02}Improved
Use output distortion instead of AP to learn the taxonomy (ablation)Unified detector · Mean of COCO, Objects365, OpenImagesmAP: 41.9{ $\pm$0.02} -> 41.7{ $\pm$0.09}Hurt
Vary the pruning threshold τ from 0.2 to 0.3 (ablation)Unified label-space learning · COCOmAP: 41.6 -> 41.8No clear effect
Retrain the detector on the unified taxonomy after weight merging (ablation)Unified detector · COCOmAP: 44.4 -> 45.4Improved
Train across four source datasets instead of using an ensemble of single-dataset modelsUnified detector · Cross-dataset meanmAP@0.5: 44.7 -> 47.3Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy