Guohao Sun's Projects
CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
Official Pytorch implementation of 'Visual Recognition with Deep Nearest Centroids'. (ICLR2023 Spotlight)
Robust vision-language understanding via evidential learning
extract feature use clip
NLP based project, including documents preprocessing using different methods, clustering and classification.
HSI reconstruction using Transformer in Pytorch
Code for "MixMatch - A Holistic Approach to Semi-Supervised Learning"
An open-source framework for training large multimodal models.
Visual self-questioning for large vision-language assistant.
SCI generation based the structure of mmcv(Video-Swin-Transformer)
Config files for my GitHub profile.