Developing Scalable CNN for Building Damage Identification
Data and Time:
March 28th (Thursday) 12 - 2 PM Central, USA.
Watch course:
Course notes:
Please view the course notes online
Description:
This webinar will guide you through building and deploying effective Convolutional Neural Networks (CNNs) for automated building damage identification. We’ll cover image classification techniques, CNN fundamentals, and hands-on experience to empower you with the skills to scale your solutions.
Key Topics:
- Understanding Image Classification and CNNs: Explore basic machine learning models, Multi-layer Perceptrons (MLPs), and the core components of CNNs for image analysis.
- The Building Blocks of CNNs: Delve into the essential elements of CNNs, including convolutional layers, pooling, activation functions, and more.
- Training Deep Learning Models: Learn strategies for effectively training your models, including data augmentation and transfer learning techniques.
- Workflow for Deep Learning: Gain practical knowledge with a step-by-step deep learning workflow.
- Case Study: Natural Hazard Detection: See real-world applications of CNNs for building damage identification after natural disasters.
- Hands-on Session: Get hands-on experience building and deploying a CNN for damage classification using PyTorch.
- Homework: Solidify your knowledge with engaging homework assignments.
Prerequisites:
Basic understanding of Python programming is recommended but not essential.
Trainer Bio:
Sikan Li is a research associate at the Texas Advanced Computing Center (TACC)’s Scalable Computational Intelligence (SCI) group. Her work focuses on developing machine learning and data mining techniques to analyze large-scale, complex datasets. She’s published several papers in this field and actively contributes to research, development, and support initiatives involving big data, statistical analysis, and machine learning at TACC. With a background in scientific data visualization, Sikan brings a unique perspective to her passion for scalable data analysis and machine learning.