Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FashionReGen: LLM-Empowered Fashion Report Generation

Yujuan Ding, Yunshan Ma, Wenqi Fan, Yige Yao, Tat-Seng Chua, Qing Li

TL;DR

The paper addresses the labor-intensive, bias-prone task of fashion report generation by introducing FashionReGen, a domain-specific benchmark and pipeline. It proposes GPT-FAR, a three-stage system that leverages catwalk-driven analysis, GPT-4V–based garment tagging, and multi-modal report generation to produce comprehensive fashion reports. Key contributions include formalizing the FashionReGen task, detailing a complete GPT-FAR pipeline, and delivering a public platform for user-generated fashion reports. The work demonstrates preliminary feasibility and highlights future opportunities to enhance data sources, report diversity, and automation for industrial-scale fashion analytics.

Abstract

Fashion analysis refers to the process of examining and evaluating trends, styles, and elements within the fashion industry to understand and interpret its current state, generating fashion reports. It is traditionally performed by fashion professionals based on their expertise and experience, which requires high labour cost and may also produce biased results for relying heavily on a small group of people. In this paper, to tackle the Fashion Report Generation (FashionReGen) task, we propose an intelligent Fashion Analyzing and Reporting system based the advanced Large Language Models (LLMs), debbed as GPT-FAR. Specifically, it tries to deliver FashionReGen based on effective catwalk analysis, which is equipped with several key procedures, namely, catwalk understanding, collective organization and analysis, and report generation. By posing and exploring such an open-ended, complex and domain-specific task of FashionReGen, it is able to test the general capability of LLMs in fashion domain. It also inspires the explorations of more high-level tasks with industrial significance in other domains. Video illustration and more materials of GPT-FAR can be found in https://github.com/CompFashion/FashionReGen.

FashionReGen: LLM-Empowered Fashion Report Generation

TL;DR

The paper addresses the labor-intensive, bias-prone task of fashion report generation by introducing FashionReGen, a domain-specific benchmark and pipeline. It proposes GPT-FAR, a three-stage system that leverages catwalk-driven analysis, GPT-4V–based garment tagging, and multi-modal report generation to produce comprehensive fashion reports. Key contributions include formalizing the FashionReGen task, detailing a complete GPT-FAR pipeline, and delivering a public platform for user-generated fashion reports. The work demonstrates preliminary feasibility and highlights future opportunities to enhance data sources, report diversity, and automation for industrial-scale fashion analytics.

Abstract

Fashion analysis refers to the process of examining and evaluating trends, styles, and elements within the fashion industry to understand and interpret its current state, generating fashion reports. It is traditionally performed by fashion professionals based on their expertise and experience, which requires high labour cost and may also produce biased results for relying heavily on a small group of people. In this paper, to tackle the Fashion Report Generation (FashionReGen) task, we propose an intelligent Fashion Analyzing and Reporting system based the advanced Large Language Models (LLMs), debbed as GPT-FAR. Specifically, it tries to deliver FashionReGen based on effective catwalk analysis, which is equipped with several key procedures, namely, catwalk understanding, collective organization and analysis, and report generation. By posing and exploring such an open-ended, complex and domain-specific task of FashionReGen, it is able to test the general capability of LLMs in fashion domain. It also inspires the explorations of more high-level tasks with industrial significance in other domains. Video illustration and more materials of GPT-FAR can be found in https://github.com/CompFashion/FashionReGen.
Paper Structure (11 sections, 1 equation, 3 figures, 1 table)

This paper contains 11 sections, 1 equation, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. GPT-FAR System
  3. Catwalk Understanding
  4. Garment tagging with GPT-4V
  5. GPT-produced tag cleaning
  6. Collective Organization and Analysis
  7. Fashion Report Generation
  8. Results and Analysis: A Case Study
  9. Report Type and Structure
  10. Generation Results and Discussion
  11. Conclusion

Figures (3)

  • Figure 1: Diagram of our GPT-FAR system.
  • Figure 2: Data flow of the catwalk understanding process.
  • Figure 3: Illustration of the Fashion Analyzing and Reporting system and an example of the generated report.