AI Paper Atlas

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

Inverse Optimal Transport for Efficient Adaptation of Vision-Language Models

Shupeng Qiu, Chuan-Xian Ren

AAAI 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Open-vocabulary alignment from VLMs · 0 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 IOTA instead of CLIP zero-shotCLIP-RN50 / ViT-B/16 · 11 fine-grained datasetsaccuracy (%): 56.04 -> 64.64 (RN50); 63.87 -> 71.73 (ViT-B/16)Improved
use IOTA instead of graph-based TransCLIP/ECALPCLIP-RN50 · 11 fine-grained datasetsaccuracy (%): ECALP 64.00 -> IOTA 64.64Improved
use IOTA with uniform prior on imbalanced Caltech101CLIP-RN50 · Caltech101accuracy (%): 85.88 -> 76.84Hurt
use IOTA instead of ECALP on domain generalizationCLIP-RN50 · ImageNet variantsaverage accuracy (%): 47.73 -> 48.43Improved
add prototype adaptation to UOT (ablation)IOTA-ZS · 11 fine-grained datasetsaccuracy (%): 61.64 -> 63.32 (+1.68)Improved
add adaptive threshold filtering (ablation)IOTA-ZS · 11 fine-grained datasetsaccuracy (%): 63.32 -> 64.32 (+1.0)Improved
add label-guided mask in few-shot setting (ablation)IOTA-FS · 16-shot, 11 datasetsaccuracy (%): 65.25 -> 70.88 (+5.63)Improved
use image-side-only marginal constraint instead of double constraint (ablation)IOTA prototype learning · ImageNetaccuracy / runtime: accuracy 64.26 -> 64.25; runtime 56.76s -> 8.93sNo clear effect
vary UOT regularization weight lambda from 0.05 to 5 (ablation)IOTA · fine-grained datasetsaverage accuracy (%): 64.25 to 64.37No clear effect
plug IOTA into CoOp prompt tuningCoOp · 11 datasets, 1-16 shotaccuracy (%): up to +4.77%Improved
plug IOTA into CLAP adapterCLAP · 11 datasets, 1-16 shotaccuracy (%): up to +2.63%Improved

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