AI-Enhanced GIS Data for Tygron: Boosting Simulation Accuracy

Course Content

1. Introduction to Image Recognition and RCNN Technology

• Objective: Understand the fundamentals of Regional Convolutional Neural Networks (RCNNs) and their application in urban planning.
• Content: Overview of image recognition principles, key concepts in convolution and pooling, and how RCNNs can identify and extract spatial features within Tygron.

2. Creating a Basic Training Dataset

• Objective: Learn how to compile and prepare image datasets for model training.
• Content: Step-by-step guide on building datasets, including creating bounding boxes around objects (e.g., trees, vegetation) for accurate identification and localization in images.

3. Training an RCNN Model with PyTorch

• Objective: Develop hands-on skills in model training, focusing on geospatial applications.
• Content: Introduction to PyTorch for RCNN model training, including data import, convolution, pooling, and key parameters affecting training accuracy. Participants will gain an understanding of how the model learns and identifies object features.

4. Saving and Integrating the Model into Tygron

• Objective: Deploy the trained model within Tygron for real-world applications.
• Content: Instructions on saving the trained model in ONNX format, compatible with Tygron, and importing it into the Tygron platform. The course will conclude with using the model for object inference, allowing identification of public space features such as trees or solar panels.

Prerequisites

Participants should have a basic understanding of Python programming and be familiar with the Tygron platform. Knowledge of AI concepts not essential.

Post-Course Support

Participants will have ongoing access to step-by-step tutorials and documentation on the Tygron Wiki, covering model implementation within Tygron. Video demonstrations will also be available to guide users on using the ONNX file for inference.

Learning Outcomes

By the end of this training, participants will be able to:

• Compile and prepare image recognition datasets tailored to geospatial and urban planning projects.
• Train an RCNN model in PyTorch for object recognition relevant to Tygron.
• Export and integrate the trained model in ONNX format into the Tygron environment.
• Leverage AI-driven image recognition for enriched data inputs and improved simulation accuracy.

This course is ideal for those looking to unlock advanced AI capabilities within the Tygron Platform, enabling more powerful insights and data-driven decision-making in urban planning projects. Contact us to secure your place in the AI-Enhanced GIS Data for Tygron course!