In the chaos of natural disasters, it's difficult for first responders to find people behind all the rubble.
This was built using Python, GPT4-V, a OpenAi's clip, Python, and Figma. We get and store disaster images and locations in a database that can then be searched to help responders find critical locations and save more lives.
Using Microsoft’s explainable GPT-V API system and a lightweight object detection system for fast processing, we can find and recover humans and animals in potentially dangerous environments such as landslides, earthquakes, and floods. By using a combination of methods, we can reduce response time in case cloud services take longer than on-device deployments. The explainability aspect of GPT4-V and understandable images is paramount to reasoning through recovery viability and methods. We aim to speed up recovery after natural disasters around the world using advanced drone technology and state-of-the-art AI algorithms.
Project Video: https://www.youtube.com/watch?v=BWdXzPxWl7U
