AI Paper Atlas

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

Subjects and Their Objects: Localizing Interactees for a Person-Centric View of Importance

Chao-Yeh Chen, Kristen Lorraine Grauman

IJCV 2016 · Detection and segmentation › Open-set and few-shot detection · approach: Human-object interaction detection · 8 citations (OpenAlex)

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 changedWhereMeasuredEffect
add interaction-guided fine-tuned CNN features (ablation)Ours-embedding · Interactee localization (COCO)IOU: 0.1780 -> 0.1989Improved
add interaction-guided fine-tuned CNN features (ablation)Ours-embedding · Interactee localization (PASCAL)IOU: 0.1998 -> 0.2177Improved
add interaction-guided fine-tuned CNN features (ablation)Ours-embedding · Interactee localization (SUN)Position error: 0.2453 -> 0.2524Mixed
replace MDN with data-driven embedding regressioninteractee predictor · Interactee localization (COCO)IOU: 0.1153 -> 0.1989Improved
use embedding instead of MDN on tiny datasetinteractee predictor · Interactee localization (SUN)Position error: 0.2356 -> 0.2524Hurt
predict interactee instead of Near Person heuristicinteractee predictor · Interactee localization (COCO)Position error: 0.2909 -> 0.2256Improved
use MDN instead of Near Person heuristicOurs-MDN · Interactee localization (COCO)IOU: 0.1213 -> 0.1153Hurt
prune DPM detections outside the predicted interactee boxDPM detector · PASCAL (computer, reading materials)detection accuracyImproved
use predicted interactee to rank important objects instead of Importance methodobject importance · COCOMention rate (%): 65.4 -> 70.5Improved
use ground truth interactee rather than predictedobject importance · COCOMention rate (%): 70.5 -> 78.4Improved
add interactee-conditioned retrieval instead of Global Matchingsentence retrieval · COCOCombined BLEU: 16.01 -> 26.43Improved
add local Content Matching to Global Matchingsentence retrieval · COCOCombined BLEU: 16.01 -> 15.20Hurt

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