davidtellez/contrastive-predictive-coding

This project helps machine learning researchers explore and implement a specific type of unsupervised learning called Contrastive Predictive Coding (CPC). It takes sequences of raw, unlabeled data, like images or sensor readings, and transforms them into meaningful representations. These learned representations can then be used to improve performance on other tasks, such as classifying images or predicting future events. Researchers and data scientists working on representation learning would find this useful.

557 stars. No commits in the last 6 months.