LLMs-based Augmentation for Domain Adaptation in Long-tailed Food Datasets

TL;DR

This paper introduces a framework using large language models to improve food recognition by addressing domain shift, long-tailed distribution, and subtle visual variations, achieving superior performance on food datasets.

Contribution

The novel framework leverages LLMs to generate descriptive texts and align them with images in a shared space for enhanced food recognition in challenging conditions.

Findings

Outperforms existing methods on long-tailed food datasets

Effective in domain adaptation for food recognition

Improves fine-grained classification accuracy

Abstract

Training a model for food recognition is challenging because the training samples, which are typically crawled from the Internet, are visually different from the pictures captured by users in the free-living environment. In addition to this domain-shift problem, the real-world food datasets tend to be long-tailed distributed and some dishes of different categories exhibit subtle variations that are difficult to distinguish visually. In this paper, we present a framework empowered with large language models (LLMs) to address these challenges in food recognition. We first leverage LLMs to parse food images to generate food titles and ingredients. Then, we project the generated texts and food images from different domains to a shared embedding space to maximize the pair similarities. Finally, we take the aligned features of both modalities for recognition. With this simple framework, we show that our proposed approach can outperform the existing approaches tailored for long-tailed data distribution, domain adaptation, and fine-grained classification, respectively, on two food datasets.