Live data from Hacker News

Tesla auto-pilot keeps confusing moon with traffic light then slowing down

reddit.com

61–70 of 93 posts

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#61
post #18

This is a problem with learning the appropriate context. In the case of a yellow traffic light, there should be a traffic light fixture and two other (unlit) lights. Without this context, it becomes apparent the yellow pixels in a circular shape (representation to the network) are not a traffic light at all. Tesla’s vision/ml systems are amazing. I would love to learn more about how unit testing for this type of erro…

> In the case of a yellow traffic light, there should be a traffic light fixture and two other (unlit) lights. Without this context, it becomes apparent the yellow pixels in a circular shape (representation to the network) are not a traffic light at all.

Permanent yield signals are often only one flashing yellow light. Crosswalk signals are often only flashing yellow lights when active. Temporary construction barrier signals are often only yellow flashing lights. Fire station signals often have only red and yellow lights, no green. Even when all three lights are present, they may also be oriented horizontally, or in triangular shapes.

Which isn't to say there isn't more context to learn from, but just about the only true unifying trait among all these indicators that you should perhaps pay attention and slow down is a bit of yellow light. It need not even be circular: yellow arrows are far from uncommon, including "straight ahead" yellow arrows for intersections where turns are forbidden.

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#62

Earlier quoted context omitted.

Which of the above would solve this problem?

LIDAR. You'd be able to tell there's no object anywhere near the car and disregard the light.

https://youtu.be/NSDTZQdo6H8?t=391

In the above video Andrej Karpathy claims that LIDAR required previously compiled maps, the creation of which is not scalable. Do you disagree with that point of view?

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#63
post #19

Earlier quoted context omitted.

Pure cost cutting at expense of safety certainly.

Humans do fine with only 2 cameras

In addition to the sibling comments, humans are also terrible drivers. They almost universally confuse confidence with ability.

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#64
post #19
post #16

I can't understand why instead of using all the possible senses: vision, auditory, radar, lidar they want to only use vision as if there's some sense of "purity" in doing that.

Pure cost cutting at expense of safety certainly.

So they could put PS5-level gaming rigs in their cars? How logical https://www.theverge.com/2021/6/1/22462660/amd-tesla-model-x...

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#65

Earlier quoted context omitted.

LIDAR. You'd be able to tell there's no object anywhere near the car and disregard the light.

Unless it's hanging from a cable https://www.masterfile.com/image/en/700-08743685 . But then, it could be a streetlight so only vision can solve this problem https://www.123rf.com/photo_2753272_hanging-street-lamp.html

You need vision to determine it is or isn't a traffic light.

Lidar would be able to tell you if the object should even be evaluated.

A light hanging from a cable would still get a lidar ping back. The moon would be a void with no lidar ping.

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#66
post #16

I can't understand why instead of using all the possible senses: vision, auditory, radar, lidar they want to only use vision as if there's some sense of "purity" in doing that.

I think one of the main reasons for Tesla to use only vision is that then the mistakes the system makes will be more understandable and thus acceptable to humans. The way radar and lidar systems fail and cause wrecks will seem strange to humans, like running into stationary objects. Even if using these sensors would make system better overall, they might not be worth including if one is shooting for human acceptance of the system as fit for use.

An autocar seeing something that looks like a red light and breaking (or breaking for a bit and then continuing after deciding it was not a red light) is going to be acceptable to people because that is how they behave. Phantom breaking due to some radar or lidar input that the human can't perceive is going to be interpreted very negatively. I imaging auditory might be added at some point, if it is not already being used.

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#67

Earlier quoted context omitted.

Radar and Lidar perhaps, being based on reflection of radio waves/laser off surfaces.

Radar isn't sensitive enough. LIDAR is.

LIDAR is so sensitive it can't reliably detect traffic lights in the rains. So you must rely on vision anyway (which is Tesla's).

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#68
post #62

Earlier quoted context omitted.

LIDAR. You'd be able to tell there's no object anywhere near the car and disregard the light.

https://youtu.be/NSDTZQdo6H8?t=391 In the above video Andrej Karpathy claims that LIDAR required previously compiled maps, the creation of which is not scalable. Do you disagree with that point of view?

Andrej's claim isn't that you need those maps, it's that using those maps is how today's LIDAR approach work.

Without the maps, you'd need to rely pretty heavily on vision to do the right thing.

What LIDAR gives you is a very accurate point cloud to work from. You can get a point cloud from vision (and indeed, that's what telsa is doing) but it won't be near as accurate as LIDAR is by default.

Re: Tesla auto-pilot keeps confusing moon with traffic light then slowing down

#70
Several European countries have a white or yellow reflective border around traffic lights to help with recognition, especially for those who might be color-blind: with a bright rectangle, it's easier to see the position of which light is lit, rather that only rely on the color.

A couple decades later, I'm now noticing some traffic lights in the US starting to gain this feature. I'm sure it will help with machines trying to parse the traffic light state... and also humans.

Post reply on HN