MarcusMaximus
Active Member
You’re missing the point. Lidar is not the key. It’s also not about hardware. Timelines are important. Inference is the most important.
Inference? Of what?
Anyway, I don’t put much stock in what the old guard thinks is necessary. Until a few years ago, all the top “experts” were sure the best ways to accomplish image recognition were with hand-made features and that neural networks would never outperform them. That lasted right up until the first time a CNN blew them out of the water. More recently, that’s been happening with image processing. People tend to be skeptical that a new technology can render all of their decades of expertise irrelevant, until it decidedly does.
Last edited: