Difficulty-Based Sampling for Debiased Contrastive Representation Learning
Authors:
- 1 June 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 24039-24048
Abstract
Contrastive learning is a self-supervised representation learning method that achieves milestone performance in various classification tasks. However, due to its unsupervised fashion, it suffers from the false negative sample problem: randomly drawn negative samples that are assumed to have a different label but actually have the same label as the anchor. This deteriorates the performance of contrastive learning as it contradicts the motivation of contrasting semantically similar and dissimilar pairs. This raised the attention and the importance of finding legitimate negative samples, which should be addressed by distinguishing between 1) true vs. false negatives; 2) easy vs. hard negatives. However, previous works were limited to the statistical approach to handle false negative and hard negative samples with hyperparameters tuning. In this paper, we go beyond the statistical approach and explore the connection between hard negative samples and data bias. We introduce a novel debiased contrastive learning method to explore hard negatives by relative difficulty referencing the bias amplifying counterpart. We propose triplet loss for training a biased encoder that focuses more on easy negative samples. We theoretically show that the triplet loss amplifies the bias in self-supervised representation learning. Finally, we empirically show the proposed method improves downstream classification performance.Keywords
Funding Information
- National Science Foundation (NSF) (2120200)
- IIS (1955890,2146091)
This publication has 14 references indexed in Scilit:
- Momentum Contrast for Unsupervised Visual Representation Learning2020
- One-Network Adversarial FairnessProceedings of the AAAI Conference on Artificial Intelligence, 2019
- Discovering Fair Representations in the Data Domain2019
- Learning Not to Learn: Training Deep Neural Networks With Biased Data2019
- The accuracy, fairness, and limits of predicting recidivismScience Advances, 2018
- Fair Prediction with Disparate Impact: A Study of Bias in Recidivism Prediction InstrumentsBig Data, 2017
- Deep Residual Learning for Image Recognition2016
- Deep Learning Face Attributes in the Wild2015
- FaceNet: A unified embedding for face recognition and clustering2015
- Learning a Similarity Metric Discriminatively, with Application to Face Verification2005