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

Lyft releases self-driving research dataset

medium.com

Post reply on HN
Print view View original

71–80 of 123 posts

Re: Lyft releases self-driving research dataset

#71

Post by skocznymroczny »

ayw wrote: post #38
cardigan wrote: post #13

(I work at scale) Hmm this blog post and the website doesn't mention that this dataset was mostly annotated by Scale (scale.ai), as part of a partnership with Lyft ... We're going to publish a blog post about this soon, but if anyone at Lyft is reading this, please figure out how to reasonably credit Scale since I doubt leaving out Scale completely from the announcement is in the spirit of the agreement. Scale should…

Hi, I'm the CEO of Scale.ai. This comment does not represent the company's viewpoint, and cardigan is not speaking on behalf of Scale. We are very excited to have been able to work with Lyft in open-sourcing this dataset and advancing the research community. We are also very grateful to Lyft for choosing to leverage our point cloud viewer and have credited the annotations to us on their launch page.

Someone will get fired today

Re: Lyft releases self-driving research dataset

#72

Post by friendly_chap »

skocznymroczny wrote: post #71
ayw wrote: post #38

Earlier quoted context omitted.

Hi, I'm the CEO of Scale.ai. This comment does not represent the company's viewpoint, and cardigan is not speaking on behalf of Scale. We are very excited to have been able to work with Lyft in open-sourcing this dataset and advancing the research community. We are also very grateful to Lyft for choosing to leverage our point cloud viewer and have credited the annotations to us on their launch page.

Someone will get fired today

Hopefully not. Cardigan obviously is has the company's interest in mind, even if perhaps the execution is a little flawed. Cardigan has just learnt a lot about PR and also gave a lot of free airtime to Scale.

Also hopefully Scale will use this opportunity to educate team members about situations like this.

Re: Lyft releases self-driving research dataset

#73

Post by tal8d »

mkagenius wrote: post #60
arathore wrote: post #9

Earlier quoted context omitted.

Such competitions do not usually result in a comprehensive "solution" by themselves - pushing the state-of-the-art is more common. Also the value is not going to be derived solely from the algorithm but more from its deployment to real world applications and the surrounding infrastructure to make it possible.

> pushing the state-of-the-art is more common But do not forget there will be 10s (if not 100s) of people working on this for 30 days. The man-hour this competition will use is highly disproportionate to the amount offered overall.

That is the point: https://en.wikipedia.org/wiki/Orteig_Prize

Re: Lyft releases self-driving research dataset

#74

Post by friendly_chap »

cardigan wrote: post #15
cardigan wrote: post #13

(I work at scale) Hmm this blog post and the website doesn't mention that this dataset was mostly annotated by Scale (scale.ai), as part of a partnership with Lyft ... We're going to publish a blog post about this soon, but if anyone at Lyft is reading this, please figure out how to reasonably credit Scale since I doubt leaving out Scale completely from the announcement is in the spirit of the agreement. Scale should…

Also, the viewer packaged with nuScenes was built by Steven Hao from Scale, and while it was packaged as part of nuScenes it should probably be called Scale's viewer instead of nuScenes' viewer. The original viewer in the nuscenes SDK has the Scale logo, but it looks like Lyft removed that in the fork. Maybe a bit of public shaming will fix that... Dear Lyft marketing person who wrote this: we are a data labeling com…

> Maybe a bit of public shaming will fix that...

Wow dude (or dudess), I was on your side but you are losing me there.

You seem to be pretty pissed, and I hope you are anonymous and not Steven Hao... There is little room for emotion in business. Grow up.

Re: Lyft releases self-driving research dataset

#75

Post by brianpgordon »

donkeyd wrote: post #70
brianpgordon wrote: post #68

Earlier quoted context omitted.

I would be surprised if many people here really just assumed that a pseudonymous user chatting with us in the HN comments was speaking on behalf of the company in an official capacity. I mean, obviously there are legal niceties to be observed and he should have appended the usual disclaimers, blah blah blah, but we do have common sense here right?

No, people don't have common sense. People should not post publicly on behalf of their employer without running it by a manager. This is lesson one at every major corporate introduction and I now understand why, because people don't have common sense.

I didn't say anything about whether he should or should not have spoken out about the deal. And I specifically said that common sense doesn't cut the mustard legally. But I am asserting that the damage from people supposedly assuming that he was speaking officially is speculative and likely zero.

Re: Lyft releases self-driving research dataset

#76

Post by KaiserPro »

jaimex2 wrote: post #18

It's looking more and more like everyone is just going to have to licence Tesla's FSD when its finished. They are the only ones with a broad real world data source and seem to have wisely taken the right path by not adopting LIDAR, focusing purely on passive vision.

Its not really that wise, its a large gamble.

tesla decided to not include lidar because they couldn't find a manufacturer that would make one cheap enough for them/fell out over terms. Its not a statement of vision. Its exactly the same decision that Apple dropped Flash support for the iPhone, The processor and ram were too limited to support it, adobe refused to make compromises, and it was too late to change before launch.

Firstly, Tesla is not focussing purely on passive vision, they are using radar as well. But because radar is nowhere near high resolution enough, they need vision to provide categorization.

Now Musk makes a lot of noise about avoiding lidar, thats mostly because he knows its a massive gamble. Yes, he bleats on about its power budget and cost, but using pure AI cost a whole more in RnD, plus a boat load of latency. Not to mention the massive power budget needed to run the custom silicon.

_eventually_ vision + radar will be more than enough to provide life critical level 5 autonomy. However Tesla barely provide more than level 2.

They have a number of problems to overcome, rain/bug occlusion of vision sensor, low light performance, sunrise/sunset, fog, reliable realtime depth estimation, etc.

I suspect that CCD based time of flight depth sensors will become cheaper, low power, and small (They are almost certainly going to end up in mobile phones soon) before pure vision realtime life critical depth estimation is a thing.

Re: Lyft releases self-driving research dataset

#77

Post by choppaface »

Just some context here:

* The raw data in nuscenes ( https://www.nuscenes.org/ ) is about 5x larger than this dataset from Lyft. 300GB train vs 60GB train. Argoverse ( https://www.argoverse.org/ ) is is about 3.x larger at 200GB. The Waymo dataset will (allegedly) be an order of magnitude larger than nuscenes ( https://i2.wp.com/syncedreview.com/wp-content/uploads/2019/0... ). BDD100k ( https://bair.berkeley.edu/blog/2018/05/30/bdd/ ) is the "largest" public dataset to date, but lacks lidar, and labels are inconsistent; most of the 100,000 scenes only have one labeled frame.

* The Lyft sensor suite has bumper-mounted lidar, which is absent from other existing datasets. Point cloud data in these areas is critical for pedestrians, bikes, and various road hazards. So this dataset alone is useful for validating work trained through other means.

* The current Lyft Level 5 release has no explicit test / validation set, which is crucial for properly measuring performance of any experiment one might do with the data. In nuscenes and Argoverse, there's a small snippet dataset that helps you prepare your pipeline. Feels like Lyft might have rushed things a little here-- they could have posted a "teaser" and then the full train and test/validation set a couple weeks later.

Great to see more public data (especially from a more modern sensor suite), plus investment into a contest with prizes.

Re: Lyft releases self-driving research dataset

#78

Post by buboard »

cardigan wrote: post #13

(I work at scale) Hmm this blog post and the website doesn't mention that this dataset was mostly annotated by Scale (scale.ai), as part of a partnership with Lyft ... We're going to publish a blog post about this soon, but if anyone at Lyft is reading this, please figure out how to reasonably credit Scale since I doubt leaving out Scale completely from the announcement is in the spirit of the agreement. Scale should…

I dont care, but god what is wrong with the replies. Let the guy speak his/er mind and stop playing game of thrones

Re: Lyft releases self-driving research dataset

#79

Post by domsnar »

preommr wrote: post #31
cardigan wrote: post #15

Earlier quoted context omitted.

Also, the viewer packaged with nuScenes was built by Steven Hao from Scale, and while it was packaged as part of nuScenes it should probably be called Scale's viewer instead of nuScenes' viewer. The original viewer in the nuscenes SDK has the Scale logo, but it looks like Lyft removed that in the fork. Maybe a bit of public shaming will fix that... Dear Lyft marketing person who wrote this: we are a data labeling com…

At first I was on kind of on your side against the other comments telling you to delete your other comment. I think its important to set the record straight if you can as early as possible. A small retraction/correction isn't guaranteed to make the frontpage of HN again. But then this comment took it into a weird turn with how fast steven can learn rock climbing (seriously, i am still kind of unsure if we're talking…

I think you're just jealous that he learned to boulder L10 in a few months from scratch AND gets to number one in video games!

Re: Lyft releases self-driving research dataset

#80

Post by mkagenius »

atarian wrote: post #65
ayw wrote: post #38

Earlier quoted context omitted.

Hi, I'm the CEO of Scale.ai. This comment does not represent the company's viewpoint, and cardigan is not speaking on behalf of Scale. We are very excited to have been able to work with Lyft in open-sourcing this dataset and advancing the research community. We are also very grateful to Lyft for choosing to leverage our point cloud viewer and have credited the annotations to us on their launch page.

This is pretty effective marketing... generate some fake controversy over some small slight and have it go viral on HN.

Plot twist: Original comment was also the CEO.
Post reply on HN