Pytorch implementation for Semantic Segmentation with multi models (Deeplabv3, Deeplabv3_plus, PSPNet, UNet, UNet_AutoEncoder, UNet_nested, R2AttUNet, AttentionUNet, RecurrentUNet,, SEGNet, CENet, DsenseASPP, RefineNet, RDFNet)

Official Pytorch Code base for "Multi-Level Global Context Cross Consistency Model for Semi-Supervised Ultrasound Image Segmentation with Diffusion Model"

A Pytorch implementation for "A Concise Model for Multi-Criteria Chinese Word Segmentation with Transformer Encoder"

PyTorch Implementation of "SEEC: Segmentation-Assisted Multi-Entropy Models for Learned Lossless Image Compression"

Boundary-constrained models for 3D abdominal multi-organ segmentation (Pytorch implementation)

PyTorch-based reimplementation of the YOLO instance segmentation model featuring a full architecture replication of Ultralytics with advanced enhancements. The model integrates C3k nested blocks, C2PSA attention modules, DFL-based bounding box refinement, and depthwise separable convolutions for efficient multi-scale detection and segmentation.

Deep Learning Pipeline: A U-Net–based system for glaucoma detection using fundus images. Tasks: • Vessel segmentation (256×256 patches ×4/image) → vessel density. • Optic disc (OD) segmentation → CDR computation. Data: ~12K multi-modal images with age, IOP, VCDR. Tech: PyTorch/Keras, Pandas, Matplotlib; supports hybrid vision+clinical models.

A PyTorch implementation of a diffusion-model + U-Net pipeline for automated multi-class segmentation of knee MRI scans. Includes dataset utilities, training/evaluation workflows, and 2D/3D inference scripts, with Docker and conda support for reproducibility.

PyTorch implementation for multivariate mixture model on cardiac segmentation from multi-source images

Pytorch implementation of a multi-task learning model for detection and segmentation of different animal breeds

This is a concise implementation of Multi-Criteria Chinese Word Segmentation using a Transformer Encoder. The project leverages Pytorch for building and training the models.

In this course, you'll use PyTorch to discover image classification, object recognition, segmentation, and image generation. You'll work with both binary and multi-class image classification models, utilize pre-trained models for deep learning tasks, and master object detection with bounding boxes.

PyTorch implementation of both vanilla segmentation networks (UNet and Attention UNet) and a novel multi-task learning (MTL) UNet. The models are trained to segment the cardiac chambers (right ventricle, left ventricle) and surrounding myocardium tissue from magnetic resonance images.

Pattern recognition course project for skin lesion classification & segmentation using HAM10000 dataset. Implements PyTorch CNN (<60M params) with evaluation metrics (accuracy, IoU, Dice). Includes data leakage prevention and optional multi-task model + Docker automation.

A 3D U-Net model for brain tumor segmentation on the UPenn GBM dataset using PyTorch. Processes multi-modal MRI scans (FLAIR, T1, T1ce, T2) for accurate tumor mask prediction with Dice loss and visualization tools.

A comprehensive AI pipeline integrating Stable Diffusion for text-to-image generation, CLIP for image analysis, and Meta's SAM2 for object segmentation. Features a RESTful API built with FastAPI and PyTorch for seamless multi-model inference.

Utilized PyTorch to develop 2 models with machine learning algorithms for early detection of 3 eye diseases. Implemented a Fully Residual U-net Neural Network for pattern identification using feature maps. Used Convolutional Neural Network CNN architectures and image segmentation. Utilized neural networks for multi-class classification.

A PyTorch-based deep learning architecture designed for multi-organ segmentation in medical imaging with built-in absence detection capabilities. The model combines U-Net-style feature extraction with anatomical attention mechanisms and presence/absence detection to handle cases where anatomical structures may be missing from scans.

This repository implements a 3D brain tumor segmentation framework using multi-modal MRI data from the BraTS dataset. A baseline 3D CNN is first developed, followed by an LSNet-inspired coarse-to-fine model that combines global context and local refinement. The project is implemented using PyTorch and MONAI.