Clustering in Contrastive Learning
Mathilde discusses the challenges of contrastive methods, particularly the impracticality of computing pairwise comparisons across large datasets. She highlights the potential of clustering to enhance learning by grouping similar images, contrasting it with traditional approaches that rely on random sampling. This innovative online clustering method aims to improve efficiency and effectiveness in training algorithms.In this clip
From this podcast
Machine Learning Street Talk (MLST)
SWaV: Unsupervised Learning of Visual Features by Contrasting Cluster Assignments (Mathilde Caron)
Related Questions