Language-only Efficient Training of Zero-shot Composed Image Retrieval

TL;DR

This paper introduces LinCIR, a language-only training framework for zero-shot composed image retrieval that uses self-supervision to achieve high performance without requiring triplet datasets, significantly reducing training time.

Contribution

LinCIR is a novel zero-shot CIR method trained solely on text data using self-masking projection, enhancing scalability and generalizability over existing approaches.

Findings

Trained LinCIR in 48 minutes with CLIP ViT-G backbone.

Achieved state-of-the-art zero-shot performance on four CIR benchmarks.

Outperformed supervised methods on FashionIQ dataset.

Abstract

Composed image retrieval (CIR) task takes a composed query of image and text, aiming to search relative images for both conditions. Conventional CIR approaches need a training dataset composed of triplets of query image, query text, and target image, which is very expensive to collect. Several recent works have worked on the zero-shot (ZS) CIR paradigm to tackle the issue without using pre-collected triplets. However, the existing ZS-CIR methods show limited backbone scalability and generalizability due to the lack of diversity of the input texts during training. We propose a novel CIR framework, only using language for its training. Our LinCIR (Language-only training for CIR) can be trained only with text datasets by a novel self-supervision named self-masking projection (SMP). We project the text latent embedding to the token embedding space and construct a new text by replacing the keyword tokens of the original text. Then, we let the new and original texts have the same latent embedding vector. With this simple strategy, LinCIR is surprisingly efficient and highly effective; LinCIR with CLIP ViT-G backbone is trained in 48 minutes and shows the best ZS-CIR performances on four different CIR benchmarks, CIRCO, GeneCIS, FashionIQ, and CIRR, even outperforming supervised method on FashionIQ. Code is available at https://github.com/navervision/lincir