3 ms·
Actually there was a presentation given by one of their Lead Data Scientists describing their ML architecture. At no point was radar mentioned. They are purely
by threeseed 7y ago
Actually there was a presentation given by one of their Lead Data Scientists describing their ML architecture. At no point was radar mentioned. They are purely relying on vision to identity cars, obstacles, traffic lights etc with dozens of models each focused on one particular 'type'.
Radar by the sounds of it is being used purely as a fallback.
The question is if the vision systems fail to recognise an obstacle at high speed is the radar long range enough to compensate in time.
Presentation:
https://slideslive.com/38917690/multitask-learning-in-the-wilderness https://slideslive.com/38917690/multitask-learning-in-the-wi...
- kjksf 7y agoActually, they did mention radar during Autonomy Investor Day (https://www.youtube.com/watch?v=Ucp0TTmvqOE https://www.youtube.com/watch?v=Ucp0TTmvqOE). Also, https://www.tesla.com/autopilot https://www.tesla.com/autopilot says: "A forward-facing radar with enhanced processing provides additional data about the world on a redundant wavelength that is able to see through heavy rain, fog, dust and even the car ahead." So it's rather bizarre that you would speculate about how they're not using radar when they explicitly say they do.
- new_realist 7y agoThere is zero chance a radar will recognize a traffic light. It is prone to false positives, to boot. Its only use is tracking large metal objects: cars and trucks, for the purpose of adaptive cruise control.
- threeseed 7y agoElon Musk or the actual Data Scientist working on it. I wonder who to trust.