With A Little Help From My Friends: Nearest-neighbor Contrastive Learning Of Visual Representations

Abstract

Self-supervised learning algorithms based on instance discrimination train encoders to be invariant to pre-defined transformations of the same instance. While most methods treat different views of the same image as positives for a contrastive loss, we are interested in using positives from other instances in the dataset. Our method, Nearest-Neighbor Contrastive Learning of visual Representations (NNCLR), samples the nearest neighbors from the dataset in the latent space, and treats them as positives. This provides more semantic variations than pre-defined transformations. We find that using the nearest-neighbor as positive in contrastive losses improves performance significantly on ImageNet classification, from 71.7% to 75.6%, outperforming previous state-of-the-art methods. On semi-supervised learning benchmarks we improve performance significantly when only 1% ImageNet labels are available, from 53.8% to 56.5%. On transfer learning benchmarks our method outperforms state-of-the-art

Stats

• citations316
• S2 citations—
• github stars0
• HF likes0
• heat score18.76
• arxiv keydwibedi2021with

Related papers

• Pnnclr: Stochastic Pseudo Neighborhoods For Contrastive Learning Based Unsupervised Representation Learning Problems (2023)7.16
• Nearest-neighbor Inter-intra Contrastive Learning From Unlabeled Videos (2023)0.00
• Unsupervised Feature Learning Via Non-parametric Instance-level Discrimination (2018)25.66
• Revisiting Contrastive Methods For Unsupervised Learning Of Visual Representations (2021)3.91
• Contrastive Learning Of Visual-semantic Embeddings (2021)0.00
• Beyond Supervised Vs. Unsupervised: Representative Benchmarking And Analysis Of Image Representation Learning (2022)8.35
• Constrained Mean Shift For Representation Learning (2021)0.00
• QK Iteration: A Self-supervised Representation Learning Algorithm For Image Similarity (2021)0.00