This problem is actually pretty obvious to fix and you don't even need a neural net - it can definitely be done "old school" computer vision.
Simply detect the presence of "flashing emergency lights" in the oncoming lane and disable autopilot when present. No object detector needed. The signal is so strong it's literally flashing extremely brightly in a regular, predictable pattern - any vision grad student should be able to figure this out.
Could there be false positives? Yep, but very few things will flash quite like that, and most half baked vision engineers can do this. The worst case is literally simply the driver taking over on occasion at night (maybe near strip clubs? lol)
I don't think the problem is as simple as that. Sure flashy lights with reflectors confuse the current model sure, but can we consider that an edge-case and work from there? Also LED's flash too in a regular pattern too. I don't know though, seems like something that would have been accounted for as it's a pretty important part of driving awareness...maybe they didn't think people would want to stay on "full auto" when a police/Emergency vehicle is in the vicinity since you might have to do some sort of maneuver to get out of it's way. So yeah, I'll say it's an edge case for now.
Disclaimer: I'm an idiot so take my view with a boeing 747 load of salt
Whilst I agree with your point, I'd also like to see an additional strategy employed.
How about emergency vehicles broadcast on 5Ghz their intended route (for the next 300 meters) or just that they are blue light in the area.
This would only be active in a blue light situation. Manufacturers can then add a detection and warn the driver.
A while back I heard about car to car communications, but I heard next to nothing these days, this is a great use case if you ask me. If someone has more details of recent developments, I'd love to read more.
> This problem is actually pretty obvious to fix and you don't even need a neural net
So can be controlled collisions as such.
Simplest FMCW radars are more than enough for every emergency braking system on the market.
It's not uncommon to even use them in parallel with some more "brainy" radar imagers like on Mercedes cars. That is to make a collision preventable in case the main imaging computer hangs, or crashes.
The solution of disabling autopilot in the presence of flashing emergency lights seems so incredibly obvious to me, that I can only assume I'm missing some glaring reason for it to not have been the case since day one.
I wholeheartedly agree. It's by definition an insult and I don't believe an academic measurement of stupidity (or intelligence) is a terribly fruitful endeavor. "My math says you're stupid!" is a place some people may go with this.
I try to think in terms of bias: "what biases do I possess that may cause me to yield an outcome that is negative?" or "what bias does person X have that lead them to those actions or words?"
I feel it promotes empathy, questioning and understanding. Not name calling. At least, it's helped me to figure out some surprising things about myself and others!
What, the elongated, shiny, rotating thing that was able to accelerate without any effect on its rate of rotation? Totally a comet, definitely, 100% guaranteed.
TF's deprecation velocity was way too high for my taste. Things we wrote would stop working randomly with their updates. I feel very similar to you about the models being "buried too deep" in their (ever-changing) machine. I much preferred how easy it was to hack Caffe V1 (once you got past the funky names, etc).
These days, I really like mxnet. Torch was a disaster, but Pytorch is much better. It's not bad in production, definitely my #2.
Interesting, however this just reflects an overfitted model. Fundamentally, this photo is not that different from a fashion influencer's posts.
We're a fashion engine and our system fully detected both of the people and all of their apparel (the person in question's shirt, his pants and it sees the "printout" as a low confidence handbag, as well as his shoes).
Not to say it would be impossible to trick our system, however, this method would not be sufficient given a good object hierarchy. Our system would have to have a triple miss across two methods - would need to miss his pants and his shirt and his body with the localizer, as well as his pants and shirt with the segmenter. And, if we were serious about detecting hiding people, you'd be surprised how gosh darn reliable the shoe detector portion is.
I don't see it being terribly feasible (and definitely not reliably so). Let's just say, it's not even close at all at this point. We miss zero of these things today.
I didn't quite understand, what is the problem with Instagram? It's too popular? Or just that it's not dropping in popularity vs Facebook (due to their recent adjustments)?
I wouldn't consider Instagram to be a messaging app - in fact they're spinning off another app from Instagram just to do messaging (because it's not exactly great at that).
It sounds to me more like Facebook might have some problem and Instagram is just doing it's own thing (as are Facebook's other services).
Somehow I completely missed the news about the standalone Instagram messaging app (called Direct, for anybody interested). Looks like I should jump ship just like I did with Facebook before it's too late.
That's definitely possible they've already more or less decided. Either way, they're pulling off an impressive optimization to obtain maximum taxpayer funded subsidies.
That's incredible. Humalog is a quite similar low price in Canada. It's strange that it'd actually be economical for many American diabetics to just book a (yearly) flight to Canada and buy 12 months worth of insulin.
I fully agree that big data has essentially been renamed to AI. Which makes sense because AI today ~= ML/DL which require lots of data.
And, as someone who went through the data warehousing fad in the late 90s, there's a lot of naive belief in pouring in a lot of data and magic happening.
That said, there has been a lot of advance that, once it's happened, we just don't call it AI any longer. Route optimization (Google Maps), predictive analytics in some domains, image recognition. Yeah, a lot of it is just fuzzy pattern recognition but some of it is pretty good.
The more fundamental question IMO is how far DL can even take you. We've actually seen a lot of progress there but we also haven't seen a lot of forward motion in cognitive science for example. So do we just run out of steam in some of the areas, like autonomous vehicles, where we think we're doing pretty well today.
Big data is not AI. Big data is processing of a giant datasets with quite classic models, such as Logistic Regression in a distributed way. That's what frameworks such as Spark and Hadoop do. AI or deep learning is different. Usually, they don't distribute across many machines so well.
Simply detect the presence of "flashing emergency lights" in the oncoming lane and disable autopilot when present. No object detector needed. The signal is so strong it's literally flashing extremely brightly in a regular, predictable pattern - any vision grad student should be able to figure this out.
Could there be false positives? Yep, but very few things will flash quite like that, and most half baked vision engineers can do this. The worst case is literally simply the driver taking over on occasion at night (maybe near strip clubs? lol)