Date: August 30, 2024
Location: Room 3315, College of Forestry, Wildlife and Environment
Time: 13:00 – 14:00
Title: Automatic detection of wildlife animals from UAV thermal imagery using a You-Look-Twice (YLT) deep learning model
Attendees: 20 people
On August 30, 2024, Dr. Fang Qiu delivered an insightful presentation titled “Automatic detection of wildlife animals from UAV thermal imagery using a You-Look-Twice (YLT) deep learning model”. The seminar, attended by 20 people in person, and additional participants via Zoom, focused on groundbreaking AI applications in wildlife conservation.
Dr. Qiu introduced how to employ AI to detect wild animals in the Chitwan National Park of Nepal. The You-Look-Twice (YLT) deep learning model was central to his discussion. This model is designed to accurately detect the presence of wildlife by processing and analyzing the thermal images captured by drones. With rigorous labeling, training, validation, and classification processes, the YLT model achieved remarkable accuracy, with detection and species identification rates exceeding 97%.
The seminar concluded with an engaging Q & A session, where faculty and students asked several questions such as: How to recognize species through AI? How to solve the overfitting and underfitting problems in the predicting model? How to reduce data noise?
Dr. Qiu’s presentation highlighted the powerful intersection of AI and environmental science, sparking a lively discussion on the future of wildlife monitoring and conservation. After the seminar and Q&A session, Dr. Qiu went to Dr. Li An’s CHES lab and interacted with faculty and CFWE students.