roytalman Goto Github PK

deep_contrastive_embedding

Deep supervised conistrastive learning for small datasets (few shot learning). This repository takes labeled embedding data ,that could be extracted from pre-trained NLP, vision, or any other algorithm that extract embedding, and use deep FFN to learn new embedding that is fine-tuned for the current data. Th algorithm can improve classification per

expirament

lstm-fcn-pytorch

Pytorch implementation for "LSTM Fully Convolutional Networks for Time Series Classification"

models

Models and examples built with TensorFlow

spectral_hrv

Calculate frequency domain Heart rate variability for matlab

tripletloss

CNN Triplet Loss function implementation for matlab