F.A.C.T - Fire Aware Cloud Technology

Deep Learning

ENBO’s F.A.C.T. (Fire Aware Cloud Technology) is built up with Deep Learning methods fed up with millions of training datasets collected from the natural world in the past twenty years.

1. The manually selected features doesn’t cover all the essential representations of the fire & smoke objects in the physical environment, working effectiveness only to a few specific scenarios. In actual world, they often introduce poor adaptability and false alerts;

2. Less capability for self-improvements when the dataset increase. With the continuous development and deployment of forest fire video surveillance systems, the increasing amount of video data will offer more complex, stable, and reliable AI models.

3. Hardware constrains. Every algorithms were tuned up to bundle with vendor’s cameras to achieve the best computing power, algorithm acceleration would be sacrificed when to work with 3rd party service providers.