Abstract:In recent years, contrastive learning has been a highly favored method for self-supervised representation learning, which significantly improves the unsupervised training of deep image models. Self-supervised learning is a subset of unsupervised learning in which the learning process is supervised by creating pseudolabels from the data themselves. Using supervised final adjustments after unsupervised pretraining is one way to take the most valuable information from a vast collection of unlabeled data and teach from a small number of labeled instances. This study aims firstly to compare contrastive learning with other traditional learning models; secondly to demonstrate by experimental studies the superiority of contrastive learning during classification; thirdly to fine-tune performance using pretrained models and appropriate hyperparameter selection; and finally to address the challenge of using contrastive learning techniques to produce data representations with semantic meaning that are independent of irrelevant factors like position, lighting, and background. Relying on contrastive techniques, the model efficiently captures meaningful representations by discerning similarities and differences between modified copies of the same image. The proposed strategy, involving unsupervised pretraining followed by supervised fine-tuning, improves the robustness, accuracy, and knowledge extraction of deep image models. The results show that even with a modest 5% of data labeled, the semisupervised model achieves an accuracy of 57.72%. However, the use of supervised learning with a contrastive approach and careful hyperparameter tuning increases accuracy to 85.43%. Further adjustment of the hyperparameters resulted in an excellent accuracy of 88.70%.

Overfitting In Contrastive Learning?

OssCSE: Overcoming Surface Structure Bias in Contrastive Learning for Unsupervised Sentence Embedding

Testing for Overfitting

Machine Learning Students Overfit to Overfitting

Contrasting the landscape of contrastive and non-contrastive learning

An Overview of Overfitting and its Solutions

Convergence of End-to-End Training in Deep Unsupervised Contrastive Learning

Overfitting, Model Tuning, and Evaluation of Prediction Performance

Is Self-Supervised Learning More Robust Than Supervised Learning?

What Should Not Be Contrastive in Contrastive Learning

What Makes for Good Views for Contrastive Learning?

What makes for good views for contrastive learning

Towards the Generalization of Contrastive Self-Supervised Learning

Understand and Improve Contrastive Learning Methods for Visual Representation: A Review

Investigating Contrastive Pair Learning's Frontiers in Supervised, Semisupervised, and Self-Supervised Learning

Overfitting in adversarially robust deep learning

From Pretext to Purpose: Batch-Adaptive Self-Supervised Learning

Understanding the Behaviour of Contrastive Loss

The Power of Contrast for Feature Learning: A Theoretical Analysis

How does Contrastive Learning Organize Images?

Improved contrastive learning model via identification of false‐negatives in self‐supervised learning