Graduate student Dilara Vefa, Master of Science in data science, will be defending her thesis titled: Synthetic Image Generation from real-world data: case study Orange County 911 Calls.
- Date: Tuesday, April 23
- Time: 11 a.m.-noon
- Location: BARC 2221
Abstract:
This work addresses the challenge of insufficient image data for emergency vehicles like ambulances in the context of autonomous driving AI models. To ensure the synthetic image data created is closely related to real-world scenarios, an initial step involved detailed data analysis of 911 call records obtained from the Orange County Government. The primary objective of this analysis was to understand real-life emergency situations and their characteristics, providing a solid foundation for generating realistic synthetic images. Machine learning methods were applied for data mining on these records, enabling a thorough extraction and interpretation of patterns and insights that are crucial for developing AI models capable of recognizing and responding to emergency vehicles effectively in a variety of traffic situations.
For more information, contact Dr. Shahram Taj.