4 ms·
Can somebody who understands what Dickmanns did, can you explain? Was he actually doing object detection on road elements? How did it stay in the lanes?
by dekhn 8y ago
Can somebody who understands what Dickmanns did, can you explain? Was he actually doing object detection on road elements? How did it stay in the lanes?
- blattimwind 8y agoIt didn't use GPS or any sort of pre-fabbed maps. It drove purely by eye / machine vision using four cameras. No radar or lidar.
- RealityVoid 8y agoI do believe they DID use a radar system. A HN user that worked on the project did mention something about using such a sensor.
- dekhn 8y agoI'm asking, specifically, what machine vision technology it used? In 1994 there wasn't much machine vision (that I'm aware of) that would even be able to do lane-keeping, except under incredibly idealized conditions (empty roads, no weather, good lighting). If so, I would request that whomever claimed this guy invented the self driving car temper their claim.
- jmiserez 8y agoHere is what someone on the project had to say about it: >Our vision system relied (heavily, not exclusively) on sensing prominent horizontal features symmetric about a common centerline. (Cars and trucks, especially at the time, have a lot of horizontal lines: bumpers, window top/bottom, valence, etc.) Thread here: https://news.ycombinator.com/item?id=10328687 https://news.ycombinator.com/item?id=10328687
- dekhn 8y agoAh! Yes, I also watched the video and I see they are identifying little "cups" (one horizontal line and two vertical). Bridges also have many hoizontal features symmetric about a common centerline. I would expect many false positives driving near bridges.
- usernam33 8y agoThe linked video states they could recognize obstacles years later, so I think there was no reading of signs etc. involved. It must have been only lane mark recognition that was used to steer. The van from the vid came years after the two mercedes s cars. I think there was no breaking for obstacles, exiting the highway, or even using interchanges. It was simply holding a lane (maybe change it when told to) and holding a set speed. Correct me if I am wrong, as I did not read into the experiment. https://www.youtube.com/watch?v=I39sxwYKlEE https://www.youtube.com/watch?v=I39sxwYKlEE
- dekhn 8y agoOK, if this is correct, then I wouldn't call this a self-driving car, or basically it was L1. Not interesting, and frankly, takes away from the dramatic results of recent self-driving cars.
- RealityVoid 8y agoThis does not seem to be the case. Check out this guy that worked on the project: https://news.ycombinator.com/item?id=10333126 https://news.ycombinator.com/item?id=10333126 The had auto-breaking, car detection and lane merging the least.
- dekhn 8y agoThe text you linked to isn't very convincing for any of those. it describes a false positive that nearly caused an injury to the head of the project. I think it's safe to say that while this was pioneering work, none of the technology it used was acceptable in terms of safety or generalizability.
- jmiserez 8y agoWhether today’s approaches are acceptable in these terms also remains to be seen.
- RealityVoid 8y ago