Real-time semantic image segmentation on mobile devices

Semantic image segmentation network with pyramid atrous convolution and boundary-aware loss for Tensorflow.

A basic Android application that runs semantic image segmentation using Tensorflow-Lite

This GitHub repository presents a scalable and reproducible framework for utilising Mapillary data in greenness visibility modelling.

[IROS 2021] ADD: A Fine-grained Dynamic Inference Architecture for Semantic Image Segmentation

Some Papers on Semantic Image Segmentation

Tutorial reimplementation of Monteiro et al. (2020) on a toy problem.

This project implements semantic image segmentation using two popular convolutional neural network architectures: U-Net and SegNet. Semantic image segmentation involves partitioning an image into multiple segments, each representing a different class.

Application of U-Net to semantic image segmentation of road images by pixel-wise classification