SimO Loss: Anchor-Free Contrastive Loss for Fine-Grained Supervised Contrastive Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bouhsine, Taha, Aaroussi, Imad El, Faysal, Atik, Huaxia, Wang
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917796324048896
author Bouhsine, Taha
Aaroussi, Imad El
Faysal, Atik
Huaxia, Wang
author_facet Bouhsine, Taha
Aaroussi, Imad El
Faysal, Atik
Huaxia, Wang
contents We introduce a novel anchor-free contrastive learning (AFCL) method leveraging our proposed Similarity-Orthogonality (SimO) loss. Our approach minimizes a semi-metric discriminative loss function that simultaneously optimizes two key objectives: reducing the distance and orthogonality between embeddings of similar inputs while maximizing these metrics for dissimilar inputs, facilitating more fine-grained contrastive learning. The AFCL method, powered by SimO loss, creates a fiber bundle topological structure in the embedding space, forming class-specific, internally cohesive yet orthogonal neighborhoods. We validate the efficacy of our method on the CIFAR-10 dataset, providing visualizations that demonstrate the impact of SimO loss on the embedding space. Our results illustrate the formation of distinct, orthogonal class neighborhoods, showcasing the method's ability to create well-structured embeddings that balance class separation with intra-class variability. This work opens new avenues for understanding and leveraging the geometric properties of learned representations in various machine learning tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2410_05233
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SimO Loss: Anchor-Free Contrastive Loss for Fine-Grained Supervised Contrastive Learning
Bouhsine, Taha
Aaroussi, Imad El
Faysal, Atik
Huaxia, Wang
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
We introduce a novel anchor-free contrastive learning (AFCL) method leveraging our proposed Similarity-Orthogonality (SimO) loss. Our approach minimizes a semi-metric discriminative loss function that simultaneously optimizes two key objectives: reducing the distance and orthogonality between embeddings of similar inputs while maximizing these metrics for dissimilar inputs, facilitating more fine-grained contrastive learning. The AFCL method, powered by SimO loss, creates a fiber bundle topological structure in the embedding space, forming class-specific, internally cohesive yet orthogonal neighborhoods. We validate the efficacy of our method on the CIFAR-10 dataset, providing visualizations that demonstrate the impact of SimO loss on the embedding space. Our results illustrate the formation of distinct, orthogonal class neighborhoods, showcasing the method's ability to create well-structured embeddings that balance class separation with intra-class variability. This work opens new avenues for understanding and leveraging the geometric properties of learned representations in various machine learning tasks.
title SimO Loss: Anchor-Free Contrastive Loss for Fine-Grained Supervised Contrastive Learning
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.05233