Official Pytorch Code base for "MobileUtr: Revisiting the relationship between light-weight CNN and Transformer for efficient medical image segmentation"

3D virtual staining with 2D and 2.5D U-Nets

3D cerebrovascular volume segmentation in Pytorch.

An example of easytorch implementation on retinal vessel segmentation.

Model training code for "A seasonally invariant deep transform for visual terrain-relative navigation"

Entries for the 2023 5th National College Student Integrated Circuit EDA Elite Challenge. SoC chip physical layout static IR drop prediction project based on methods such as image processing and NLP unsupervised learning.

U-Net

Monte Carlo dropout method for uncertainty quantification in image segmentation

Segmentation of cancerous tumors using Mamba. Code, resources, and paper provided. We manage to make a small (42k param) model that can segment pretty well.

Transforming 2D images into 3D semantically segmented scenes using innovative CNN architecture and COLMAP reconstruction.

Train a U-net model for pixel-wise segmentation of facial wrinkles. Based on FFHQ dataset.

Land cover classification in Tanzania using ensemble labels and high resolution Planet NICFI basemaps and Sentinel-1 time series.

Utilizing U-NET deep-learning to deconvolve Structured Illumination Microscopy (SIM) Images. A clean and concise python implemenation.

This research work basically highlights my undergrad thesis works. In my thesis, I have worked on the BraTS 2020 dataset. My total journey of thesis from building various models to writing paper is presented here.

crystal structure analysis with RHEED and ML

MICCAI2019: 3D U2-Net: A 3D Universal U-Net for Multi-Domain Medical Image Segmentation

Implementation of U-net and pipeline of changing color on human hair.

Modular PyTorch U-Net model

基于ResNet改进U-Net，融合注意力机制和ASPP模块，实现多类别语义分割。Improve U-Net based on ResNet, integrate attention mechanism and ASPP module, and realize multi-category semantic segmentation.

Project outside of course scope at (BSc) Machine Learning and Data Science education programme. Colab between NGI and DIKU at University of Copenhagen.