AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

TIPSv2: Advancing Vision-Language Pretraining with Enhanced Patch-Text Alignment

Bingyi Cao, Koert Chen, Kevis-Kokitsi Maninis, Kaifeng Chen, Arjun Karpur, Ye Xia, Sahil Dua, Tanmaya Dabral, Guangxing Han, Bohyung Han, Joshua Ainslie, Alex Bewley and 7 more

CVPR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Recaptioning and text-supervised pretraining · 10 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
replace iBOT with iBOT++ (loss on visible tokens too) (ablation)TIPS ViT-g · ADE150 zero-shot segmIoU: 3.5 -> 17.6Improved
replace iBOT with iBOT++ (ablation)TIPS ViT-g · PC59 zero-shot segmIoU: 14.2 -> 28.6Improved
replace iBOT with iBOT++ (ablation)TIPS ViT-g · PASCAL linear-probe segmIoU: 82.8 -> 82.5No clear effect
replace iBOT with iBOT++ (ablation)TIPS ViT-g · ImageNet KNNaccuracy: 83.2 -> 84.4Improved
add multi-granularity (Gemini + PaliGemma) captions (ablation)TIPSv2 ViT-g · Flickr T2I retrievalRecall@1: 81.7 -> 85.4Improved
add multi-granularity captions (ablation)TIPSv2 ViT-g · ADE150 zero-shot segmIoU: 17.6 -> 18.1Improved
add head-only EMA (ablation)TIPSv2 ViT-g · ADE150 zero-shot segmIoU: 18.1 -> 19.1Improved
add head-only EMA (ablation)TIPSv2 ViT-g · ImageNet KNNaccuracy: 84.3 -> 84.1No clear effect
raise iBOT++ masking ratio from 0.0 to 0.75 (ablation)TIPS ViT-L · ADE150 zero-shot segmIoU: 1.0 -> 13.6Improved
use two CLS tokens with Gemini captions instead of one CLS (web, PaliGemma) (ablation)TIPSv2 ViT-g · COCO I2T retrievalRecall@1: 72.3 -> 76.2Improved
add iBOT++ instead of iBOT (ablation)CLIP ViT-L · PC60 zero-shot segmIoU: 8.0 -> 22.9Improved
distill ViT-L student from ViT-g teacherTIPS · ADE150 zero-shot segmIoU: 2.6 -> 20.8Improved

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