Search results

He mentions that they dont use many segemntation tasks but I dont quite understand that. It seem like the whole birds eye view is based on semantic segmentation, lane lines, curbes, dividers and so on. Could anyone help me shed a light on this, what am I misunderstanding?

Interested in this too. Didnt some fans buy a board in new york or something a while back? Perhaps this could be the same case?

I dont know if this was posted before, but for anyone wanna read more about this here is source. https://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(20)30154-9.pdf

I am not 100 but I do believe that Gali, hyperchange, did comment on the rides. Something like "The bounding boxes were so tight", I am probably misquoting. I dont remember if he said anything else about the ride though.