Enhancing Zero-Shot Facial Expression Recognition by LLM Knowledge Transfer

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhao, Zengqun, Cao, Yu, Gong, Shaogang, Patras, Ioannis
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909404871262208
author Zhao, Zengqun
Cao, Yu
Gong, Shaogang
Patras, Ioannis
author_facet Zhao, Zengqun
Cao, Yu
Gong, Shaogang
Patras, Ioannis
contents Current facial expression recognition (FER) models are often designed in a supervised learning manner and thus are constrained by the lack of large-scale facial expression images with high-quality annotations. Consequently, these models often fail to generalize well, performing poorly on unseen images in inference. Vision-language-based zero-shot models demonstrate a promising potential for addressing such challenges. However, these models lack task-specific knowledge and therefore are not optimized for the nuances of recognizing facial expressions. To bridge this gap, this work proposes a novel method, Exp-CLIP, to enhance zero-shot FER by transferring the task knowledge from large language models (LLMs). Specifically, based on the pre-trained vision-language encoders, we incorporate a projection head designed to map the initial joint vision-language space into a space that captures representations of facial actions. To train this projection head for subsequent zero-shot predictions, we propose to align the projected visual representations with task-specific semantic meanings derived from the LLM encoder, and the text instruction-based strategy is employed to customize the LLM knowledge. Given unlabelled facial data and efficient training of the projection head, Exp-CLIP achieves superior zero-shot results to the CLIP models and several other large vision-language models (LVLMs) on seven in-the-wild FER datasets. The code and pre-trained models are available at https://github.com/zengqunzhao/Exp-CLIP.
format Preprint
id arxiv_https___arxiv_org_abs_2405_19100
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Enhancing Zero-Shot Facial Expression Recognition by LLM Knowledge Transfer
Zhao, Zengqun
Cao, Yu
Gong, Shaogang
Patras, Ioannis
Computer Vision and Pattern Recognition
Current facial expression recognition (FER) models are often designed in a supervised learning manner and thus are constrained by the lack of large-scale facial expression images with high-quality annotations. Consequently, these models often fail to generalize well, performing poorly on unseen images in inference. Vision-language-based zero-shot models demonstrate a promising potential for addressing such challenges. However, these models lack task-specific knowledge and therefore are not optimized for the nuances of recognizing facial expressions. To bridge this gap, this work proposes a novel method, Exp-CLIP, to enhance zero-shot FER by transferring the task knowledge from large language models (LLMs). Specifically, based on the pre-trained vision-language encoders, we incorporate a projection head designed to map the initial joint vision-language space into a space that captures representations of facial actions. To train this projection head for subsequent zero-shot predictions, we propose to align the projected visual representations with task-specific semantic meanings derived from the LLM encoder, and the text instruction-based strategy is employed to customize the LLM knowledge. Given unlabelled facial data and efficient training of the projection head, Exp-CLIP achieves superior zero-shot results to the CLIP models and several other large vision-language models (LVLMs) on seven in-the-wild FER datasets. The code and pre-trained models are available at https://github.com/zengqunzhao/Exp-CLIP.
title Enhancing Zero-Shot Facial Expression Recognition by LLM Knowledge Transfer
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2405.19100