News for hackers. Stuff that matters.
Re:Oyster: Towards Unsupervised Object Detection from Lidar ... by williamcotton on Tuesday June 13, 2023 @08:11AM (#36309073)
No, I can't, because there's a lot more going on than just having recognized an object. How can we tell that it is cardboard? Or if we see someone throw a ball into the street, how can we get an idea of how much mass the ball has? Or any number of such scenarios... the answer is, we have a lifetime of experience of interacting with our environment. We've seen cardboard boxes, felt them, crushed them up and tossed them around. We've thrown balloons and rolled bowling balls. We have a learned understanding of both ourselves and the objects in our environment and how they behave.
Re:Oyster: Towards Unsupervised Object Detection from Lidar ... by the__alchemist on Tuesday June 13, 2023 @08:22AM (#36309201)
> No, I can't, because there's a lot more going on than just having recognized an object. How can we tell that it is cardboard

How can my ears tell that an object is red? Eyes are fantastic things.

Re:Oyster: Towards Unsupervised Object Detection from Lidar ... by williamcotton on Tuesday June 13, 2023 @08:26AM (#36309227)
Knowing the color, shape, or sound that something makes doesn't mean anything without a lifetime of experiences to frame those sensory inputs and especially with regards to driving!

You're being obtuse!

Re:Oyster: Towards Unsupervised Object Detection from Lidar ... by the__alchemist on Tuesday June 13, 2023 @08:47AM (#36309457)
I've been dancing around the point in the hope you'd come to your own conclusions. Removing the last trace of subtlety:

Your main point is this: > How does LIDAR help with [things cameras are good at and LIDAR is not]?

This is a strawman, because no one claims it does. LIDAR's value is as one of multiple sensors to fuse together to build a coherant model.

Re:Oyster: Towards Unsupervised Object Detection from Lidar ... by williamcotton on Tuesday June 13, 2023 @08:58AM (#36309602)
That's not my point, and I was explicit about this in my original comment! My point is that sensory input to an autonomous driving system is not the difficult part to solve! Just take a quick look at some of my other comments so I don't have to repeat myself.