SemanticSegmentation

Image Semantic Segmentation with different models (e.g. DeeplabV3+ and HRNetV2) on different datasets (e.g. Cityscapes, Camvid, BDD100k, IDD)

Pre-requisites

• Python 3.6
• PyTroch 1.5
• Numpy 1.18
• Scipy 1.6.3
• Scikit-learn 0.23.1
• Opencv-python 4.2

Dataset

The code uses the following datasets:

1. Cityscapes
2. Camvid
3. BDD100k
4. Indian Driving Dataset (IDD)

Pretrained models

• The pretrained DeeplabV3+ on BDD100K dataset model
• The pretrained DeeplabV3+ on Indian Driving Dataset (IDD) dataset model

Contact

Razieh Kaviani Baghbaderani (rkavian1@vols.utk.edu)