AI Paper Atlas

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

Contrastive Feature Masking Open-Vocabulary Vision Transformer

Dahun Kim, Anelia Angelova, Weicheng Kuo

ICCV 2023 · Detection and segmentation › Open-set and few-shot detection · approach: Open-vocabulary object detection · 28 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 masked feature reconstruction in joint image-text embedding space (ablation)contrastive image-text pretraining · LVIS OVDAPr: 27.4 -> 30.7Improved
reconstruct raw pixels or first-layer features instead of final features (ablation)CFM-ViT · LVIS OVDAPr: 27.4 -> 27.1 (pixels), 27.2 (1st-layer)No clear effect
add positional embedding dropout in pretraining (ablation)contrastive baseline · LVIS OVDAPr: 27.4 -> 28.5Improved
PED plus masked feature reconstruction (ablation)CFM-ViT · LVIS OVDAPr: 27.4 -> 31.2Improved
remove positional embeddings entirely instead of dropout (ablation)ViT encoder · LVIS OVDAPr: 27.4 -> 25.8Hurt
mask contrastive branch to 75% tokens, disjoint from reconstruction input (ablation)CFM-ViT · LVIS OVDAPr and FLOPs: FLOPs 1.23x -> 1.01x; APr 30.7 -> 30.4Improved
use backbone lr ratio 0.5 instead of 0.0, 0.1 or 1.0 (ablation)detection finetuning · LVIS OVDAPr: 26.0 -> 27.4Improved
frozen backbone inference for region classification, with PED pretraining (ablation)CFM-ViT · LVIS OVDAPr: 31.2 -> 32.5Improved
frozen backbone inference without PED pretraining (ablation)CFM-ViT · LVIS OVDAPr: 30.7 -> 27.1Hurt
swap pretraining data from ALIGN to LAION (ablation)CFM-ViT · LVIS OVDAPr: 32.5 -> 32.4 (LAION-2B), 32.2 (LAION-400M)No clear effect
add feature reconstruction plus PED (ablation)contrastive baseline · Flickr30K zero-shot retrievalimage-to-text R@1: 86.0 -> 87.0Improved
increase batch size from 4k to 16k (ablation)CFM-ViT · LVIS OVDAPr: ViT-B 26.8 -> 28.8; ViT-L 32.5 -> 33.9Improved

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