Semantic_Segmentation_Using_RESNET18_Encoder_FPN_Decoder

This repository includes the code to do END-TO-END semantic segmentation of road scenses using Camvid dataset.

Note:

• The resulting model obtained is not the optimal model.It has been trained just for few epochs.
• The emphasis is laid on End-to End semantic segmenation pipeline (DataLoaders,and Training pipiline) rather than results.
• Feelfree to start from the checkpoint and fine-tune the model.

TO RUN

1.Clone this repository:

git clone --recurse-submodules https://github.com/sairampolina/Roadscenes_Semantic_Segmentation_FPN_Decoder.git

2.Setup a conda environment and Install Dependencies

 conda create --name <env_name> --file requirements.txt

3.To auto-download data and plot loss and metric curves RUN (from root of the directory)

python3 experiment.py

4.For inference on few test samples RUN:

python3 test.py