AI Paper Atlas

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

Wukong: A 100 Million Large-scale Chinese Cross-modal Pre-training Benchmark

Jiaxi Gu, Xiaojun Meng, Guansong Lu, Lu Hou, Minzhe Niu, Xiaodan Liang, Lewei Yao, Runhui Huang, Wei Zhang, Xin Jiang, Chunjing Xu, Hang Xu

NeurIPS 2022 · Vision-language and multimodal › CLIP and zero-shot · approach: Self-distillation for vision-language pretraining · 29 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
use character-grained instead of word-grained Chinese tokenization (ablation)Wukong ViT-B · Zero-shot classification, 10 Chinese datasetsavg top-1 accuracy: 62.65 -> 65.65Improved
lock the image encoder (LiT-tuning) rather than train it (ablation)Wukong ViT-B · ImageNet zero-shot validation accuracyvalidation accuracyImproved
lock the image encoder (LiT-tuning) (ablation)Wukong ViT-B · training losstraining lossHurt
add token-wise similarity with reduced-token interaction (Wukong-style) vs FILIP-style on ViT-BViT-B model · Zero-shot classification, 10 Chinese datasetsavg top-1 accuracy: 58.09 -> 65.65Improved
add reduced-token interaction to Swin-L encoder (Wukong vs FILIP style)Swin-L model · Zero-shot classification, 10 Chinese datasetsavg top-1 accuracy: 70.27 -> 69.12Hurt
Wukong ViT-L (reduced-token) vs CLIP ViT-L (global similarity)ViT-L model · Zero-shot classification, 10 Chinese datasetsavg top-1 accuracy: 72.60 -> 73.03Improved
pre-train on Wukong dataset (CLIP/FILIP/Wukong models) vs BriVL on its own dataChinese VLP models · Zero-shot classification, 10 Chinese datasetsavg top-1 accuracy: 43.72 -> 73.03Improved
Wukong ViT-L vs WenLan 2.0 pre-trained on 650M pairsWukong ViT-L · AIC-ICC fine-tuned retrievalmean recall: 58.7 -> 71.6Improved
global-similarity CLIP ViT-L vs Wukong ViT-L on in-domain Wukong-TestCLIP ViT-L · Wukong-Test fine-tuned retrievalmean recall: 85.8 -> 89.6Mixed

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