An Efficient Post-hoc Framework for Reducing Task Discrepancy of Text Encoders for Composed Image Retrieval

• URL: http://arxiv.org/abs/2406.09188v2
• Date: Tue, 18 Mar 2025 04:06:55 GMT
• Title: An Efficient Post-hoc Framework for Reducing Task Discrepancy of Text Encoders for Composed Image Retrieval
• Authors: Jaeseok Byun, Seokhyeon Jeong, Wonjae Kim, Sanghyuk Chun, Taesup Moon,
• Abstract summary: Composed Image Retrieval (CIR) aims to retrieve a target image based on a reference image and conditioning text, enabling controllable image searches.<n>Traditional Zero-Shot (ZS) CIR methods bypass the need for expensive training CIR triplets by projecting image embeddings into the text token embedding space.<n>We introduce Reducing Taskrepancy of Texts (RTD), an efficient text-only framework that complements projection-based CIR methods.
• Score: 34.065449743428005
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Composed Image Retrieval (CIR) aims to retrieve a target image based on a reference image and conditioning text, enabling controllable image searches. The mainstream Zero-Shot (ZS) CIR methods bypass the need for expensive training CIR triplets by projecting image embeddings into the text token embedding space, forming a composed query for retrieval. However, we highlight an inherent limitation in these projection-based CIR: a task discrepancy of text encoders between the original pre-training task of the encoders (text $\leftrightarrow$ image) and the target CIR task (image + text $\leftrightarrow$ image), which potentially negatively impacts CIR performance. To reduce such a discrepancy, a naive solution would be to train both image and text encoders with CIR triplets in a supervised manner. Instead, we introduce Reducing Task Discrepancy of Text Encoders (RTD), an efficient text-only post-hoc framework that complements projection-based CIR methods. We devise a novel target-anchored text contrastive learning designed to enhance the capability of the text encoder for CIR. We also propose two key enhancements: (1) a hard negative-based refined batch sampling strategy and (2) a refined concatenation scheme to further mitigate training-inference discrepancy. Integrating RTD into state-of-the-art projection-based methods achieves performance comparable to, or even surpassing, resource-intensive state-of-the-art synthetic CIR triplet-based approaches only with 23 minutes of additional training on 4 A100 GPUs (up to $100\times$ faster in training). Our code will be available upon acceptance.

Related papers

• Fine-grained Textual Inversion Network for Zero-Shot Composed Image Retrieval [60.20835288280572]
  We propose a novel Fine-grained Textual Inversion Network for ZS-CIR, named FTI4CIR. FTI4CIR comprises two main components: fine-grained pseudo-word token mapping and tri-wise caption-based semantic regularization.
  arXiv  Detail & Related papers  (2025-03-25T02:51:25Z)
• PDV: Prompt Directional Vectors for Zero-shot Composed Image Retrieval [37.95145173167645]
  We introduce Prompt Directional Vector (PDV), a simple yet effective training-free enhancement that captures semantic modifications induced by user prompts. PDV enables three key improvements: (1) dynamic composed text embeddings where prompt adjustments are controllable via a scaling factor, (2) composed image embeddings through semantic transfer from text prompts to image features, and (3) weighted fusion of composed text and image embeddings.
  arXiv  Detail & Related papers  (2025-02-11T03:20:21Z)
• MoTaDual: Modality-Task Dual Alignment for Enhanced Zero-shot Composed Image Retrieval [20.612534837883892]
  Composed Image Retrieval (CIR) is a challenging vision-language task, utilizing bi-modal (image+text) queries to retrieve target images. In this paper, we propose a two-stage framework to tackle both discrepancies. MoTaDual achieves the state-of-the-art performance across four widely used ZS-CIR benchmarks, while maintaining low training time and computational cost.
  arXiv  Detail & Related papers  (2024-10-31T08:49:05Z)
• Decoder Pre-Training with only Text for Scene Text Recognition [54.93037783663204]
  Scene text recognition (STR) pre-training methods have achieved remarkable progress, primarily relying on synthetic datasets. We introduce a novel method named Decoder Pre-training with only text for STR (DPTR) DPTR treats text embeddings produced by the CLIP text encoder as pseudo visual embeddings and uses them to pre-train the decoder.
  arXiv  Detail & Related papers  (2024-08-11T06:36:42Z)
• HyCIR: Boosting Zero-Shot Composed Image Retrieval with Synthetic Labels [5.34016463729574]
  Composed Image Retrieval (CIR) aims to retrieve images based on a query image with text. Current Zero-Shot CIR (ZS-CIR) methods try to solve CIR tasks without using expensive triplet-labeled training datasets. We propose Hybrid CIR (HyCIR), which uses synthetic labels to boost the performance of ZS-CIR.
  arXiv  Detail & Related papers  (2024-07-08T09:55:36Z)
• Spherical Linear Interpolation and Text-Anchoring for Zero-shot Composed Image Retrieval [43.47770490199544]
  Composed Image Retrieval (CIR) is a complex task that retrieves images using a query, which is configured with an image and a caption. We introduce a novel ZS-CIR method that uses Spherical Linear Interpolation (Slerp) to directly merge image and text representations. We also introduce Text-Anchored-Tuning (TAT), a method that fine-tunes the image encoder while keeping the text encoder fixed.
  arXiv  Detail & Related papers  (2024-05-01T15:19:54Z)
• Leveraging Cross-Modal Neighbor Representation for Improved CLIP Classification [54.96876797812238]
  We present a novel CrOss-moDal nEighbor Representation(CODER) based on the distance structure between images and their neighbor texts. The key to construct a high-quality CODER lies in how to create a vast amount of high-quality and diverse texts to match with images. Experiment results across various datasets and models confirm CODER's effectiveness.
  arXiv  Detail & Related papers  (2024-04-27T02:04:36Z)
• Visual Delta Generator with Large Multi-modal Models for Semi-supervised Composed Image Retrieval [50.72924579220149]
  Composed Image Retrieval (CIR) is a task that retrieves images similar to a query, based on a provided textual modification. Current techniques rely on supervised learning for CIR models using labeled triplets of the reference image, text, target image. We propose a new semi-supervised CIR approach where we search for a reference and its related target images in auxiliary data.
  arXiv  Detail & Related papers  (2024-04-23T21:00:22Z)
• Expediting Contrastive Language-Image Pretraining via Self-distilled Encoders [10.649402840032138]
  ECLIPSE features a distinctive distillation architecture wherein a shared text encoder is utilized between an online image encoder and a momentum image encoder. Based on the unified text embedding space, ECLIPSE compensates for the additional computational cost of the momentum image encoder by expediting the online image encoder.
  arXiv  Detail & Related papers  (2023-12-19T23:11:06Z)
• Training-free Zero-shot Composed Image Retrieval with Local Concept Reranking [34.31345844296072]
  Composed image retrieval attempts to retrieve an image of interest from gallery images through a composed query of a reference image and its corresponding modified text. Most current composed image retrieval methods follow a supervised learning approach to training on a costly triplet dataset composed of a reference image, modified text, and a corresponding target image. We present a new training-free zero-shot composed image retrieval method which translates the query into explicit human-understandable text.
  arXiv  Detail & Related papers  (2023-12-14T13:31:01Z)
• Language-only Efficient Training of Zero-shot Composed Image Retrieval [46.93446891158521]
  Composed image retrieval (CIR) task takes a composed query of image and text, aiming to search relative images for both conditions. 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)
  arXiv  Detail & Related papers  (2023-12-04T16:22:06Z)
• Pretrain like Your Inference: Masked Tuning Improves Zero-Shot Composed Image Retrieval [17.70430913227593]
  We introduce a novel unlabeled and pre-trained masked tuning approach to reduce the gap between the pre-trained model and the downstream CIR task. With such a simple design, it can learn to capture fine-grained text-guided modifications.
  arXiv  Detail & Related papers  (2023-11-13T02:49:57Z)
• Vision-by-Language for Training-Free Compositional Image Retrieval [78.60509831598745]
  Compositional Image Retrieval (CIR) aims to retrieve the relevant target image in a database. Recent research sidesteps this need by using large-scale vision-language models (VLMs) We propose to tackle CIR in a training-free manner via Vision-by-Language (CIReVL)
  arXiv  Detail & Related papers  (2023-10-13T17:59:38Z)
• Sentence-level Prompts Benefit Composed Image Retrieval [69.78119883060006]
  Composed image retrieval (CIR) is the task of retrieving specific images by using a query that involves both a reference image and a relative caption. We propose to leverage pretrained V-L models, e.g., BLIP-2, to generate sentence-level prompts. Our proposed method performs favorably against the state-of-the-art CIR methods on the Fashion-IQ and CIRR datasets.
  arXiv  Detail & Related papers  (2023-10-09T07:31:44Z)
• Pic2Word: Mapping Pictures to Words for Zero-shot Composed Image Retrieval [84.11127588805138]
  Composed Image Retrieval (CIR) combines a query image with text to describe their intended target. Existing methods rely on supervised learning of CIR models using labeled triplets consisting of the query image, text specification, and the target image. We propose Zero-Shot Composed Image Retrieval (ZS-CIR), whose goal is to build a CIR model without requiring labeled triplets for training.
  arXiv  Detail & Related papers  (2023-02-06T19:40:04Z)
• CRIS: CLIP-Driven Referring Image Segmentation [71.56466057776086]
  We propose an end-to-end CLIP-Driven Referring Image framework (CRIS) CRIS resorts to vision-language decoding and contrastive learning for achieving the text-to-pixel alignment. Our proposed framework significantly outperforms the state-of-the-art performance without any post-processing.
  arXiv  Detail & Related papers  (2021-11-30T07:29:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.