GohVh / resnet34-unet

Segmentation model using UNET architecture with ResNet34 as encoder background, designed with PyTorch.

Resnet34-Unet

This segmentation model is an UNET architecture with ResNet34 as encoder background.

🌟 Architecture Diagram

📷 *Diagram will be uploaded later

🏃 Run

Clone the project

  git clone https://github.com/GohVh/resnet34-unet.git

Open your Jupyter notebook/Google Colab notebook

  %run main.py

  %run predict.py

🚩 Results

Loss

Score

Accuracy

Predicted results

🤝 Contact

• LinkedIn Profile: GohVh
• Email: [e-mail]

💎 Acknowledgements

The dataset used in this project is a Semantic Drone Dataset focuses on semantic understanding of urban scenes for increasing the safety of autonomous drone flight and landing procedures. This dataset is prepared by a team of members from Institute of Computer Graphics and Vision, Graz University of Technology.

• Aerial Segmentation Dataset

Below are the journal papers of two architectures used in this project:

• ResNet34 paper, Official implementation of Resnet
• Unet paper, Official implementation of Unet