Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tesla's self driving algorithm's overlay [video]

tesla.com

Post reply on HN
Print view View original

541–550 of 658 posts

Re: Tesla's self driving algorithm's overlay [video]

#541

Post by gogoincar »

virtualritz wrote: post #493
Animats wrote: post #472

Compare Google's system from five years ago.[1] (at 7:42) This new video is at n00b level compared to that. [1] https://embed.ted.com/talks/chris_urmson_how_a_driverless_ca...

The camera is bird's eye on the google presentation. And it is a recording. So jitter/flicker can be cleaned up/smoothed out and the data can be massaged in ways that a real time system may not be able to do. This is a presentation. I'd be /very/ surprised if this animation was from RAW data as-is. Also there seems to be Lidar data (point clouds) which Tesla doesn't have. So while this means bounding boxes may have l…

Those scenarios from Google were obviously selected for the presentation, but very much represent state of the art in _real time_ processing for autonomous driving systems from the time of the presentation.

I’m not sure what is implied with saying it’s a recording - both the Google and Tesla presentations are “recordings” and equal opportunity to pick best case examples, but I would bet strongly there is nothing not RAW = “real time“ for their respective compute platforms.

The top down viewpoint helps show off the quality (still by no means perfect) of the world representation. If you projected Tesla’s model into 3D you would see far more jitter than in the video overlay for a variety of reasons.

That said, I think comparing them directly on specific technical components is a bit of a sidebar. They are taking two very different paths along the way to a still ambiguous problem. Both are leading their respective approaches, but have fundamentally different and unproven assumptions.

Also worth looking not just at how accurately objects are detected but what the visualizations show about the intent of other road users. The Google video shows predicted trajectories for important objects in a number of scenes. We don’t get to see any of that clearly from Tesla, and that is by no means a small part of the problem. Not sure if it is there and not shown, just highlighting there is a lot more downstream even once you are finding objects reliably in the sensors.

Re: Tesla's self driving algorithm's overlay [video]

#542

Post by tim333 »

felipemnoa wrote: post #183
brian_ wrote: post #163

If Teslas navigate with cameras, does that mean a Wil E Coyote style tunnel painting will actually work?

If done right it may even work against humans.

It's sort of been tried https://artofgears.com/2015/12/16/did-a-fiat-crash-into-a-pa...

A problem I think is the visual appearance of a tunnel changes as you move in relationship to in in a way that a static image would not.

Re: Tesla's self driving algorithm's overlay [video]

#543

Post by Dylan16807 »

xkemp wrote: post #426
axguscbklp wrote: post #119

Earlier quoted context omitted.

>I thought fully autonomous driving was meant to be done by now? Musk was bullshitting when he made that prediction. Maybe he did it because he had bought his own bullshit. Personally I think what's more likely is that he was cynically conning people. In reality, I think no-one is anywhere close to fully self driving cars. I would be surprised if we saw fully self driving cars any time in the next 50 years. The whole…

I don't understand why "otherwise smart people" feel the need to diminish the potential and successes of ML. When chess was basically solved, it was dismissed brute force, with a sarcastic challenge to do Go, and we now the fate of that challenge. There used to be, on slashdot and I believe in the early days of HN, this running complaint of the new-at-the-time CSI-style shows, specifically the "enhance" trope: "you c…

> When chess was basically solved, it was dismissed brute force

Deep blue was brute force. There are different ways to solve a problem, and only some of them are impressive in certain ways, even if they're impressive in other ways.

> Today we have superresolution algorithms that can reconstruct license plates from low-res images.

You need to take another look at CSI. You can reconstruct a little bit, usually with the help of multiple frames or exact font data. You can't do this: https://www.youtube.com/watch?v=I_8ZH1Ggjk0 https://www.youtube.com/watch?v=3uoM5kfZIQ0

> Many tech people dismiss such progress as "just statistics", but I haven't seen much of an attempt to find a definition of intelligence that is meaningful different from "just statistics". In fact I doubt it's possible within the realm of science, i. e. without resorting to mind-body dualism.

That's about as useful as saying everything is 'just chemistry'. It's true but not helpful. And the same way that your brain and a science fair volcano are both just chemistry, demonstrating a simple reaction isn't impressive.

Re: Tesla's self driving algorithm's overlay [video]

#544

Post by Dylan16807 »

spookthesunset wrote: post #445
xkemp wrote: post #426

Earlier quoted context omitted.

I don't understand why "otherwise smart people" feel the need to diminish the potential and successes of ML. When chess was basically solved, it was dismissed brute force, with a sarcastic challenge to do Go, and we now the fate of that challenge. There used to be, on slashdot and I believe in the early days of HN, this running complaint of the new-at-the-time CSI-style shows, specifically the "enhance" trope: "you c…

They exist now as what amounts to fancy amusement park rides. Wake me up when I can get sloppy drunk and catch one home. Only I was drunk and didn’t provide the exactly correct drop off address and need to somehow communicate to the robo-Uber where exactly it should drop me off... (how will I do that, by the way? talk to it? What if it doesn’t understand my accent?) Also, I’m at a wedding and the pickup is in a grass…

Your phone knows where your home is.

What if the human doesn't understand your accent!

> Every trip that isn’t a pre-planned, set route amusement park ride is an edge case.

That is not even remotely true.

And you could have a fleet that's 80% or 90% robot. It's not hard. $5 discount if your end points aren't all wonky and robot-confusing.

Re: Tesla's self driving algorithm's overlay [video]

#545

Post by iamaelephant »

m463 wrote: post #495
virtualritz wrote: post #493

Earlier quoted context omitted.

The camera is bird's eye on the google presentation. And it is a recording. So jitter/flicker can be cleaned up/smoothed out and the data can be massaged in ways that a real time system may not be able to do. This is a presentation. I'd be /very/ surprised if this animation was from RAW data as-is. Also there seems to be Lidar data (point clouds) which Tesla doesn't have. So while this means bounding boxes may have l…

It diminishes the comparison when just the lidar for a waymo car costs more than the entire tesla vehicle. So, not only can you afford a tesla vehicle, you can go out and buy one today.

Can you go out and buy a fully self driving Tesla today? Because that would be the relevant comparison.

Re: Tesla's self driving algorithm's overlay [video]

#546

Post by titzer »

Is it just me, or other people out there terrified of trusting their life to some closed-source neural net that sees in black and white at 18fps which apparently believes that cars pop into and out of existence in a matter of milliseconds? This is not the future I hoped for.

Re: Tesla's self driving algorithm's overlay [video]

#547

Post by iamaelephant »

Veedrac wrote: post #506
Animats wrote: post #472

Compare Google's system from five years ago.[1] (at 7:42) This new video is at n00b level compared to that. [1] https://embed.ted.com/talks/chris_urmson_how_a_driverless_ca...

Note that Google's system operates in pre-mapped areas and uses lidar.

So it operates under the conditions that we expect the commercially available product to operate under? Yeah cool thanks for the caveat.

Re: Tesla's self driving algorithm's overlay [video]

#548

Post by mike_d »

eru wrote: post #501
mike_d wrote: post #468

Earlier quoted context omitted.

If that is true you should call in to question the integrity of a company that would run life-critical software on a non-RTOS.

Realtime has nothing to do with absolute performance. It's about meeting your deadlines. And, of course, often real-time systems have much worse throughput performance than a non-real-time system on the same hardware. After all, latency guarantees are not free.

Yup. I think we agree on that.

It doesn't matter if you have a highly performant pipeline for detecting other cars if you have a random 200 ms VM pause as another car blows a stop sign in front of you.

Re: Tesla's self driving algorithm's overlay [video]

#549

Post by iamaelephant »

Shivetya wrote: post #121
ArtWomb wrote: post #7

I believe this was released as part of Tesla's Autopilot hiring announcement. More details here: https://www.tesla.com/autopilotAI Definitely someone's dream job ;) I particularly like the applicant query: "Tell us, what extraordinary work you have done?"

while an interesting video I really want to see the entire front arc stitched together. How is it judging it is safe to go through the stop sign? Situations like that fascinate me the most. One issue not discussed enough is all this push for automation really needs road marking guidelines pushed down from the Federal level. While the feds can hold domain over the interstate system or roads it can be maddening the dif…

> One issue not discussed enough is all this push for automation really needs road marking guidelines pushed down from the Federal level

Yup especially because every country is USA.

Re: Tesla's self driving algorithm's overlay [video]

#550

Post by andbberger »

aeternum wrote: post #90
iamleppert wrote: post #79

Not sure I'd trust an autopilot based on crude visual odometry and segmentation that can only achieve a paltry 17 fps.

Not sure I'd trust humans since their crude visual systems lead to reaction times of over one second: https://www.researchgate.net/publication/233039156_Brake_Rea... 17x faster is an impressive improvement

Really doesn't matter how fast it is if it doesn't work does it? Don't let all the impressive CNN results fool you, we are a long, long way from doing human level vision.

There's a reason everyone else uses lidar...

Post reply on HN