Evaluating Contrastive Models For Instance-based Image Retrieval

Abstract

In this work, we evaluate contrastive models for the task of image retrieval. We hypothesise that models that are learned to encode semantic similarity among instances via discriminative learning should perform well on the task of image retrieval, where relevancy is defined in terms of instances of the same object. Through our extensive evaluation, we find that representations from models trained using contrastive methods perform on-par with (and outperforms) a pre-trained supervised baseline trained on the ImageNet labels in retrieval tasks under various configurations. This is remarkable given that the contrastive models require no explicit supervision. Thus, we conclude that these models can be used to bootstrap base models to build more robust image retrieval engines.

Stats

• citations4
• S2 citations—
• github stars0
• HF likes0
• heat score5.24
• arxiv keykrishna2021evaluating

Related papers

• Unsupervised Dense Information Retrieval With Contrastive Learning (2021)0.00
• Unsupervised Dense Retrieval With Relevance-aware Contrastive Pre-training (2023)10.44
• Unsupervised Dense Retrieval With Conterfactual Contrastive Learning (2024)0.00
• Benchmarking Robustness Of Contrastive Learning Models For Medical Image-report Retrieval (2025)0.00
• Keyword-based Diverse Image Retrieval By Semantics-aware Contrastive Learning And Transformer (2023)5.24
• Revisiting Contrastive Methods For Unsupervised Learning Of Visual Representations (2021)3.91
• Relevance-based Margin For Contrastively-trained Video Retrieval Models (2022)7.74
• With A Little Help From My Friends: Nearest-neighbor Contrastive Learning Of Visual Representations (2021)18.76