Saved in:
Bibliographic Details
Main Authors: Czerwinska, Urszula, Bircanoglu, Cenk, Chamoux, Jeremy
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2504.07567
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908311407820800
author Czerwinska, Urszula
Bircanoglu, Cenk
Chamoux, Jeremy
author_facet Czerwinska, Urszula
Bircanoglu, Cenk
Chamoux, Jeremy
contents We benchmark foundation models image embeddings for classification and retrieval in e-Commerce, evaluating their suitability for real-world applications. Our study spans embeddings from pre-trained convolutional and transformer models trained via supervised, self-supervised, and text-image contrastive learning. We assess full fine-tuning and transfer learning (top-tuning) on six diverse e-Commerce datasets: fashion, consumer goods, cars, food, and retail. Results show full fine-tuning consistently performs well, while text-image and self-supervised embeddings can match its performance with less training. While supervised embeddings remain stable across architectures, SSL and contrastive embeddings vary significantly, often benefiting from top-tuning. Top-tuning emerges as an efficient alternative to full fine-tuning, reducing computational costs. We also explore cross-tuning, noting its impact depends on dataset characteristics. Our findings offer practical guidelines for embedding selection and fine-tuning strategies, balancing efficiency and performance.
format Preprint
id arxiv_https___arxiv_org_abs_2504_07567
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Benchmarking Image Embeddings for E-Commerce: Evaluating Off-the Shelf Foundation Models, Fine-Tuning Strategies and Practical Trade-offs
Czerwinska, Urszula
Bircanoglu, Cenk
Chamoux, Jeremy
Computer Vision and Pattern Recognition
Artificial Intelligence
Computational Engineering, Finance, and Science
Information Retrieval
Machine Learning
We benchmark foundation models image embeddings for classification and retrieval in e-Commerce, evaluating their suitability for real-world applications. Our study spans embeddings from pre-trained convolutional and transformer models trained via supervised, self-supervised, and text-image contrastive learning. We assess full fine-tuning and transfer learning (top-tuning) on six diverse e-Commerce datasets: fashion, consumer goods, cars, food, and retail. Results show full fine-tuning consistently performs well, while text-image and self-supervised embeddings can match its performance with less training. While supervised embeddings remain stable across architectures, SSL and contrastive embeddings vary significantly, often benefiting from top-tuning. Top-tuning emerges as an efficient alternative to full fine-tuning, reducing computational costs. We also explore cross-tuning, noting its impact depends on dataset characteristics. Our findings offer practical guidelines for embedding selection and fine-tuning strategies, balancing efficiency and performance.
title Benchmarking Image Embeddings for E-Commerce: Evaluating Off-the Shelf Foundation Models, Fine-Tuning Strategies and Practical Trade-offs
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computational Engineering, Finance, and Science
Information Retrieval
Machine Learning
url https://arxiv.org/abs/2504.07567