LLMs generate image quality labels to boost e-commerce sales

Researchers have developed a method called Image Score to evaluate image quality for e-commerce platforms like Mercari. This approach utilizes Large Language Models (LLMs) with Chain-of-Thought prompting to generate aesthetic labels for product images. The LLM-generated labels correlate with user behavior and are more cost-effective than human evaluation, leading to improved customer journey optimization and a significant increase in sales during online experimentation. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT This LLM-driven image quality assessment could improve product discovery and sales conversion on e-commerce platforms.

RANK_REASON This is a research paper detailing a new method for image quality assessment using LLMs.

Read on arXiv cs.CV →

• Mercari
• Image Score
• LLM
• Chain-of-Thought
• arXiv

• paper
• product
• other