Talks

Wildfires are a disastrous phenomenon which cause damage to land, loss of property, air pollution, and even loss of human life. Due to the warmer and drier conditions created by climate change, more severe and uncontrollable wildfires are expected to occur in the coming years. This could lead to a global wildfire crisis and have dire consequences on our planet. Hence, it has become imperative to use technology to help prevent the spread of wildfires. One way to prevent the spread of wildfires before they become too large is to perform early detection i.e, detecting the smoke before the actual fire starts. In this paper, we present our Wildfire Detection and Alert System which use machine learning to detect wildfire smoke with a high degree of accuracy and can send immediate alerts to users. Our technology is currently being used in the USA to monitor data coming in from hundreds of cameras daily. We show that our system has a high true detection rate and a low false detection rate. Our performance evaluation study also shows that on an average our system detects wildfire smoke faster than an actual person.

Wildfires are a huge problem in the state of California. In 2021 alone, there have been 7,000+ wildfires in California, which have burned down over 2 million acres of land. An automated system that detects wildfires early, before they spread, can help save lives and minimize damage to land and structures. In this talk, we will discuss Alchera's Wildfire Alert System, which uses machine learning to perform early detection of wildfires. The system analyzes images in real-time from over 800 cameras and sends immediate alerts to users.

Talk was based on the following paper:

• Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model: https://arxiv.org/abs/1911.08265

Talk was based on the following paper:

• Applying Machine Learning to Crowd-sourced Data from Earthquake Detective: https://arxiv.org/abs/2011.04740

Talk was based on the following paper:

• Applying Machine Learning to Crowd-sourced Data from Earthquake Detective: https://arxiv.org/abs/2011.04740

Talk was based on the following papers:

• Recurrent Experience Replay in Distributed RL (R2D2): https://openreview.net/pdf?id=r1lyTjAqYX
• Curiosity Driven RL: https://arxiv.org/pdf/1705.05363.pdf
• Random Distillation Networks: https://arxiv.org/pdf/1810.12894.pdf
• Never Give Up (NGU): https://arxiv.org/pdf/2002.06038.pdf
• Agent57: https://arxiv.org/pdf/2003.13350.pdf

Talk was based on the following paper:

• OtoWorld: Towards Learning to Separate by Learning to Move:: https://arxiv.org/abs/2007.06123

Talk was based on the following papers:

• Imagination Augmented Agents: https://arxiv.org/abs/1707.06203
• Curiosity Driven Exploration: https://arxiv.org/abs/1705.05363
• Large Scale Study of Curiosity: https://arxiv.org/abs/1808.04355

The talk was based on the following papers:

• Autocurricula: https://arxiv.org/abs/1903.00742
• Emergent Tool Use (Hide and Seek): https://arxiv.org/abs/1909.07528
• Alpha Go Zero: https://www.nature.com/articles/nature24270