Corrembed: Evaluating Pre-trained Model Image Similarity Efficacy With A Novel Metric

Abstract

Detecting visually similar images is a particularly useful attribute to look to when calculating product recommendations. Embedding similarity, which utilizes pre-trained computer vision models to extract high-level image features, has demonstrated remarkable efficacy in identifying images with similar compositions. However, there is a lack of methods for evaluating the embeddings generated by these models, as conventional loss and performance metrics do not adequately capture their performance in image similarity search tasks. In this paper, we evaluate the viability of the image embeddings from numerous pre-trained computer vision models using a novel approach named CorrEmbed. Our approach computes the correlation between distances in image embeddings and distances in human-generated tag vectors. We extensively evaluate numerous pre-trained Torchvision models using this metric, revealing an intuitive relationship of linear scaling between ImageNet1k accuracy scores and tag-correlat

Stats

• citations4
• S2 citations—
• github stars0
• HF likes0
• heat score5.24
• arxiv keyborgersen2023corrembed

Related papers

• Learning Non-metric Visual Similarity For Image Retrieval (2017)11.58
• Needle In A Haystack, Fast: Benchmarking Image Perceptual Similarity Metrics At Scale (2022)0.00
• Correlation Verification For Image Retrieval (2022)18.80
• Benchmarking Pretrained Vision Embeddings For Near- And Duplicate Detection In Medical Images (2023)7.16
• Predicting Visual Overlap Of Images Through Interpretable Non-metric Box Embeddings (2020)8.82
• Effectively Leveraging Attributes For Visual Similarity (2021)3.66
• Visual Product Search Benchmark (2026)0.00
• Genecis: A Benchmark For General Conditional Image Similarity (2023)10.07