uzairakbar / efficient-segmentation

Semantic Segmentation with reduced fully convolutional networks for higher latency and lower memory requirement.

PyTorch implementation of FCNNs for Semantic Segmentation followed by Pruning for Resource Efficient Inference

This project lists all of the project deliverables for the TUM project course 'Artificial Intelligence for Embedded Systems' (Winter Semester 2018/ 2019).

Project Deliverables

• Project Report
• Project Proposal
• Status Report 1
• Status Report 2

Dataset

The MSRC v2 dataset for semantic segmentation.

Sample (test set) Results