AI Paper Atlas

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

A Unified Masked Jigsaw Puzzle Framework for Vision and Language Models

Weixin Ye, Wei Wang, Yahui Liu, Yue Song, Bin Ren, Wei Bi, Rita Cucchiara, Nicu Sebe

TPAMI 2025 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining · 2 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 MJP (shuffle + unk position embedding + DAL) to DeiT-SDeiT-S · ImageNet-1K valTop-1 accuracy: 79.8 -> 80.5Improved
add MJP to Swin-TSwin-T · ImageNet-1K valTop-1 accuracy: 81.3 -> 81.3No clear effect
add MJP to ResNet-50 (patch shuffle only)ResNet-50 · ImageNet-1K valTop-1 accuracy: 79.3 -> 79.4No clear effect
add MJP to DeiT-BDeiT-B · ImageNet-1K valTop-1 accuracy: 81.8 -> 82.0Improved
use masking ratio 0.03 instead of 0 when training DeiT-S (ablation)DeiT-S + MJP · ImageNet-1K valTop-1 accuracy: 80.0 -> 80.5Improved
add MJP to Swin-Tiny backbone for UperNetSwin-Tiny · ADE20K segmentationmIoU: 43.87 -> 44.03Improved
add MJP instead of naive shuffle to BERT-baseBERT-base · Yelp sentimentTop-1 accuracy: 70.34 -> 70.68Improved
add MJP instead of naive shuffle to BERT-baseBERT-base · Amazon sentimentTop-1 accuracy: 72.03 -> 72.17Improved
add MJP to Transformer4Transformer4 · Yelp sentimentTop-1 accuracy: 57.89 -> 58.05Improved
use n-gram window size 32 rather than 128 for MJP (ablation)BERT-base + MJP · Yelp sentimentTop-1 accuracy: 70.25 -> 70.68Mixed
add DAL/DRL auxiliary position loss to MJP for text (ablation)BERT-base + MJP · Yelp sentimentTop-1 accuracyNo clear effect
raise masking ratio to 0.7-0.8 for BERT-base on Yelp (ablation)BERT-base + MJP · Yelp sentimentTop-1 accuracy: 70.68 -> 70.07Hurt

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