AI Paper Atlas

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

Hierarchy-Aware Pseudo Word Learning with Text Adaptation for Zero-Shot Composed Image Retrieval

Zhe Li, Lei Zhang, Zheren Fu, Kun Zhang, Zhendong Mao

ICCV 2025 · Vision-language and multimodal › CLIP and zero-shot · approach: Attribute-based zero-shot learning · 6 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 object-aware tokens s_s on top of global token (ablation)HIT (BLIP) · CIRR valR@1: 33.2 -> 35.0Improved
add object-aware tokens s_p on top of global token (ablation)HIT (BLIP) · CIRR valR@1: 33.2 -> 35.7Improved
use all three levels s_g, s_s and s_p together (ablation)HIT (BLIP) · CIRR valR@1: 35.7 -> 37.3Improved
use all three levels instead of s_g plus s_s (ablation)HIT (BLIP) · Fashion-IQ valR@10: 32.9 -> 34.5Improved
add two-stage grouping modules (L=2 vs none) (ablation)HIT (BLIP) · CIRR valavg R@1+R@5: 49.08 -> 53.35Improved
increase grouping modules beyond 2 (L=2 to L=4) (ablation)HIT (BLIP) · CIRR valavg R@1+R@5: 53.35 -> 51.12Hurt
use 32 and 8 group tokens instead of 16 and 4 (ablation)HIT (BLIP) · CIRR valR@1: 34.6 -> 37.3Improved
increase group tokens from 32/8 to 64/32 (ablation)HIT (BLIP) · CIRR valR@1: 37.3 -> 34.9Hurt
increase group tokens from 32/8 to 32/16 (ablation)HIT (BLIP) · Fashion-IQ valR@10: 34.5 -> 33.8Hurt
set auxiliary loss weight to 0.7 instead of 0.1 (ablation)HIT (BLIP) · CIRR valavg R@1+R@5: 51.46 -> 53.35Improved
raise auxiliary loss weight from 0.7 to 1.0 (ablation)HIT (BLIP) · CIRR valavg R@1+R@5: 53.35 -> 52.97Hurt
use stronger BLIP-2 backbone instead of CLIP (ablation)HIT · CIRR valR@1: 28.7 -> 39.8Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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