IEEE

Harris Performance as Activation Function for Image Classification in Convolutional Neural Network and Transfer Learning Models

Abstract: Activation functions are important parameters of a deep neural network (DNN) which introduces non-linearity, complexity, and adaptability of the model for image classification. Convolutional ...
GitHub

Code release for CL-GPT (Convolutional Long-term structure with Gaussian Process Transfer)

An end-to-end scalable deep recurrent structure with fast transfer learning is proposed to accurately estimate SOC for lithium-ion batteries at different temperatures. First, convolutional and ...
IEEE

A Fast SAR Image Generation Method Based on System Transfer Function Convolution

Abstract: The scarcity of synthetic aperture radar (SAR) datasets is a fundamental issue in the field of SAR image applications. Large-scale, diverse and high-quality data is the key to achieving good ...
C&EN

CFTR_TL: Transfer Learning-Enhanced Prediction of CFTR ATP Binding Sites with Multi-Window Convolutional Neural Networks

Creative Commons (CC): This is a Creative Commons license. Attribution (BY): Credit must be given to the creator. The Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) protein, crucial for ...