SOTAVerified

Zero-Shot Composed Image Retrieval (ZS-CIR)

Given a query composed of a reference image and a relative caption, Composed Image Retrieval (CIR) aims to retrieve target images that are visually similar to the reference one but incorporate the changes specified in the relative caption. The bi-modality of the query provides users with more precise control over the characteristics of the desired image, as some features are more easily described with language, while others can be better expressed visually.

Zero-Shot Composed Image Retrieval (ZS-CIR) is a subtask of CIR that aims to design an approach that manages to combine the reference image and the relative caption without the need for supervised learning.

Papers

Showing 110 of 36 papers

TitleStatusHype
MegaPairs: Massive Data Synthesis For Universal Multimodal RetrievalCode3
MagicLens: Self-Supervised Image Retrieval with Open-Ended InstructionsCode3
An Efficient Post-hoc Framework for Reducing Task Discrepancy of Text Encoders for Composed Image RetrievalCode2
LDRE: LLM-based Divergent Reasoning and Ensemble for Zero-Shot Composed Image RetrievalCode2
Reason-before-Retrieve: One-Stage Reflective Chain-of-Thoughts for Training-Free Zero-Shot Composed Image RetrievalCode2
Semantic Editing Increment Benefits Zero-Shot Composed Image RetrievalCode2
Composed Image Retrieval for Remote SensingCode2
iSEARLE: Improving Textual Inversion for Zero-Shot Composed Image RetrievalCode2
CoVR-2: Automatic Data Construction for Composed Video RetrievalCode1
Context-I2W: Mapping Images to Context-dependent Words for Accurate Zero-Shot Composed Image RetrievalCode1
Show:102550
← PrevPage 1 of 4Next →

No leaderboard results yet.