AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

ECCV Caption: Correcting False Negatives by Collecting Machine-and-Human-Verified Image-Caption Associations for MS-COCO

Sanghyuk Chun, Wonjae Kim, Song Park, Minsuk Chang, Seong Joon Oh

ECCV 2022 · Vision-language and multimodal › Human and embodied modeling · approach: Concept-guided image captioning · 37 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
evaluate with ECCV mAP@R instead of COCO 1K R@1CLIP ViT-L/14 vs PCME · ECCV Captionranking: 55.4% vs 40.1% COCO 1K R@1 -> 28.0% vs 37.1% ECCV mAP@RMixed
use semi-hard instead of hardest negative mining (ablation)PVSE K=1 · ECCV CaptionmAP@R: 35.76 -> 36.63Improved
add hardest negative mining (ablation)PVSE K=1 · COCO 1KR@1: 43.28 -> 49.12Improved
add hardest negative mining (vs none) (ablation)PVSE K=1 · ECCV CaptionmAP@R: 33.34 -> 35.76Improved
add semi-hard negative mining (vs none) (ablation)PVSE K=1 · ECCV CaptionmAP@R: 33.34 -> 36.63Improved
increase PVSE embeddings K from 1 to 2 (ablation)PVSE · ECCV CaptionmAP@R: 33.98 -> 40.26Improved
add CutMix pre-training (ablation)PCME · ECCV CaptionmAP@R: 37.11 -> 41.74Improved
fine-tune VinVL on COCO Caption (ablation)VinVL · ECCV CaptionmAP@R: 22.18 -> 40.81Improved
fine-tune ViLT on COCO Caption (ablation)ViLT · ECCV CaptionmAP@R: 26.84 -> 34.58Improved
use COCO 1K R@1 ranking to predict ECCV mAP@R ranking25 VL models · rank correlationKendall tau: 0.47Hurt
use PMRP ranking to predict ECCV mAP@R ranking25 VL models · rank correlationKendall tau: 0.20Hurt
replace R@1 with mAP@R as metric in human studymetric · human preference alignmentpreference: A: 70.85, B: 10.66, C: 13.15, D: 4.89, E: 0.44Improved

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