English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Composed Image Retrieval (CIR) is a challenging multimodal task thatretrieves a target image based on a reference image and accompanyingmodification text. Due to the high cost of annotating CIR triplet datasets,zero-shot (ZS) CIR has gained traction as a promising alternative. Existingstudies mainly focus on projection-based methods, which map an image to asingle pseudo-word token. However, these methods face three criticalchallenges: (1) insufficient pseudo-word token representation capacity, (2)discrepancies between training and inference phases, and (3) reliance onlarge-scale synthetic data. To address these issues, we propose a two-stageframework where the training is accomplished from mapping to composing. In thefirst stage, we enhance image-to-pseudo-word token learning by introducing avisual semantic injection module and a soft text alignment objective, enablingthe token to capture richer and fine-grained image information. In the secondstage, we optimize the text encoder using a small amount of synthetic tripletdata, enabling it to effectively extract compositional semantics by combiningpseudo-word tokens with modification text for accurate target image retrieval.The strong visual-to-pseudo mapping established in the first stage provides asolid foundation for the second stage, making our approach compatible with bothhigh- and low-quality synthetic data, and capable of achieving significantperformance gains with only a small amount of synthetic data. Extensiveexperiments were conducted on three public datasets, achieving superiorperformance compared to existing approaches.

From Mapping to Composing: A Two-Stage Framework for Zero-shot Composed Image Retrieval