Saved in:
Bibliographic Details
Main Authors: Pham, Chau, Phan, Hoang, Doermann, David, Tian, Yunjie
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.17610
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912167417085952
author Pham, Chau
Phan, Hoang
Doermann, David
Tian, Yunjie
author_facet Pham, Chau
Phan, Hoang
Doermann, David
Tian, Yunjie
contents The personalization model has gained significant attention in image generation yet remains underexplored for large vision-language models (LVLMs). Beyond generic ones, with personalization, LVLMs handle interactive dialogues using referential concepts (e.g., ``Mike and Susan are talking.'') instead of the generic form (e.g., ``a boy and a girl are talking.''), making the conversation more customizable and referentially friendly. In addition, PLVM is equipped to continuously add new concepts during a dialogue without incurring additional costs, which significantly enhances the practicality. PLVM proposes Aligner, a pre-trained visual encoder to align referential concepts with the queried images. During the dialogues, it extracts features of reference images with these corresponding concepts and recognizes them in the queried image, enabling personalization. We note that the computational cost and parameter count of the Aligner are negligible within the entire framework. With comprehensive qualitative and quantitative analyses, we reveal the effectiveness and superiority of PLVM.
format Preprint
id arxiv_https___arxiv_org_abs_2412_17610
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Personalized Large Vision-Language Models
Pham, Chau
Phan, Hoang
Doermann, David
Tian, Yunjie
Computer Vision and Pattern Recognition
The personalization model has gained significant attention in image generation yet remains underexplored for large vision-language models (LVLMs). Beyond generic ones, with personalization, LVLMs handle interactive dialogues using referential concepts (e.g., ``Mike and Susan are talking.'') instead of the generic form (e.g., ``a boy and a girl are talking.''), making the conversation more customizable and referentially friendly. In addition, PLVM is equipped to continuously add new concepts during a dialogue without incurring additional costs, which significantly enhances the practicality. PLVM proposes Aligner, a pre-trained visual encoder to align referential concepts with the queried images. During the dialogues, it extracts features of reference images with these corresponding concepts and recognizes them in the queried image, enabling personalization. We note that the computational cost and parameter count of the Aligner are negligible within the entire framework. With comprehensive qualitative and quantitative analyses, we reveal the effectiveness and superiority of PLVM.
title Personalized Large Vision-Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.17610