GitHub

Self-Training for Sample-Efficient Active Learning for Text Classification with Pre-Trained Language Models

Experiment code for the paper Self-Training for Sample-Efficient Active Learning for Text Classification with Pre-Trained Language Models. Accepted at EMNLP 2024 Main. Note: For the sake of any easier ...
IEEE

Sample-Level Improved Cross-Source Contrastive Learning for PAN and MS Joint Classification

Abstract: In recent years, the number and ways of acquiring panchromatic (PAN) images and multispectral (MS) images have increased, and manual labeling costs have also increased. Processing these data ...