Conceptual Codebook Learning for Vision-Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Yi, Yu, Ke, Wu, Siqi, He, Zhihai
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909255032897536
author Zhang, Yi
Yu, Ke
Wu, Siqi
He, Zhihai
author_facet Zhang, Yi
Yu, Ke
Wu, Siqi
He, Zhihai
contents In this paper, we propose Conceptual Codebook Learning (CoCoLe), a novel fine-tuning method for vision-language models (VLMs) to address the challenge of improving the generalization capability of VLMs while fine-tuning them on downstream tasks in a few-shot setting. We recognize that visual concepts, such as textures, shapes, and colors are naturally transferable across domains and play a crucial role in generalization tasks. Motivated by this interesting finding, we learn a conceptual codebook consisting of visual concepts as keys and conceptual prompts as values, which serves as a link between the image encoder's outputs and the text encoder's inputs. Specifically, for a given image, we leverage the codebook to identify the most relevant conceptual prompts associated with the class embeddings to perform the classification. Additionally, we incorporate a handcrafted concept cache as a regularization to alleviate the overfitting issues in low-shot scenarios. We observe that this conceptual codebook learning method is able to achieve enhanced alignment between visual and linguistic modalities. Extensive experimental results demonstrate that our CoCoLe method remarkably outperforms the existing state-of-the-art methods across various evaluation settings, including base-to-new generalization, cross-dataset evaluation, and domain generalization tasks. Detailed ablation studies further confirm the efficacy of each component in CoCoLe.
format Preprint
id arxiv_https___arxiv_org_abs_2407_02350
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Conceptual Codebook Learning for Vision-Language Models
Zhang, Yi
Yu, Ke
Wu, Siqi
He, Zhihai
Computer Vision and Pattern Recognition
In this paper, we propose Conceptual Codebook Learning (CoCoLe), a novel fine-tuning method for vision-language models (VLMs) to address the challenge of improving the generalization capability of VLMs while fine-tuning them on downstream tasks in a few-shot setting. We recognize that visual concepts, such as textures, shapes, and colors are naturally transferable across domains and play a crucial role in generalization tasks. Motivated by this interesting finding, we learn a conceptual codebook consisting of visual concepts as keys and conceptual prompts as values, which serves as a link between the image encoder's outputs and the text encoder's inputs. Specifically, for a given image, we leverage the codebook to identify the most relevant conceptual prompts associated with the class embeddings to perform the classification. Additionally, we incorporate a handcrafted concept cache as a regularization to alleviate the overfitting issues in low-shot scenarios. We observe that this conceptual codebook learning method is able to achieve enhanced alignment between visual and linguistic modalities. Extensive experimental results demonstrate that our CoCoLe method remarkably outperforms the existing state-of-the-art methods across various evaluation settings, including base-to-new generalization, cross-dataset evaluation, and domain generalization tasks. Detailed ablation studies further confirm the efficacy of each component in CoCoLe.
title Conceptual Codebook Learning for Vision-Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.02350