AI Paper Atlas

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

Task-Oriented Multi-Modal Mutual Leaning for Vision-Language Models

Sifan Long, Zhen Zhao, Junkun Yuan, Zichang Tan, Jiangjiang Liu, Luping Zhou, Shengsheng Wang, Jingdong Wang

ICCV 2023 · Vision-language and multimodal › CLIP and zero-shot · approach: Prompt learning for CLIP

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 class-aware text prompts (CTP) to the text branch (ablation)CoOp baseline · Base-to-new, 11-dataset averageharmonic mean accuracy (%): 71.66 -> 76.64Improved
add text-guided feature tuning (TFT) to the image branch (ablation)CoOp baseline · Base-to-new, 11-dataset averageharmonic mean accuracy (%): 71.66 -> 77.24Improved
combine CTP and TFT (ablation)CoOp baseline · Base-to-new, 11-dataset averageharmonic mean accuracy (%): 71.66 -> 79.02Improved
use text-image attention (CTP) instead of Linear-ReLU-Linear block for prompt learning (ablation)text branch · Base-to-new, 11-dataset averageharmonic mean accuracy (%): 75.83 -> 76.64Improved
use text-image attention (TFT) instead of Linear-ReLU-Linear block for feature tuning (ablation)image branch · Base-to-new, 11-dataset averageharmonic mean accuracy (%): 75.94 -> 77.24Improved
inject image information into prompts of all classes identically (CoCoOp) instead of class-aware promptsCoOp · Base-to-new, base classes, 11-dataset averageaccuracy (%): 82.69 -> 80.47Hurt
replace CoCoOp with CTP + TFTCoCoOp · Base-to-new, new classes, 11-dataset averageaccuracy (%): 71.69 -> 75.72Improved
replace CoCoOp with CTP + TFT in 16-shot learningCoCoOp · 16-shot, 11-dataset averageaccuracy (%): 74.79 -> 80.32Improved
replace CoCoOp with CTP + TFT in cross-dataset transferCoCoOp · Cross-dataset transfer, target averageaccuracy (%): 65.74 -> 66.47Improved
vary the balance weight lambda (ablation)Ours · Base-to-newaccuracyNo clear effect

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