Abstract: Street view (SV) images provide valuable supplementary data for characterizing the functional attributes of land use types, improving urban land use classification based on ...
A production-ready deep learning project for time-series image classification using EfficientNet/NFNet with PyTorch Lightning. This project implements transfer learning for multi-class classification ...
Abstract: Vision transformers (ViTs) and convolutional neural networks (CNNs) have demonstrated remarkable performance in classifying complicated hyperspectral images (HSIs). However, these models ...