Contrastive Learning via Local Activity
Contrastive learning (CL) helps deep networks discriminate between positive and negative pairs in learning. As a powerful unsupervised pretraining method, CL has greatly reduced the performance gap with supervised training. However, current CL approaches mainly rely on sophisticated augmentations, a...
Main Authors: | He Zhu, Yang Chen, Guyue Hu, Shan Yu |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-12-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/12/1/147 |
Similar Items
-
Combining Contrastive Learning with Auto-Encoder for Out-of-Distribution Detection
by: Dawei Luo, et al.
Published: (2023-12-01) -
Grouped Contrastive Learning of Self-Supervised Sentence Representation
by: Qian Wang, et al.
Published: (2023-08-01) -
DeepDiffusion: Unsupervised Learning of Retrieval-Adapted Representations via Diffusion-Based Ranking on Latent Feature Manifold
by: Takahiko Furuya, et al.
Published: (2022-01-01) -
Hierarchical skeleton Meta-Prototype Contrastive learning with hard skeleton mining for unsupervised person re-identification
by: Rao, Haocong, et al.
Published: (2023) -
Graph Clustering with High-Order Contrastive Learning
by: Wang Li, et al.
Published: (2023-10-01)