zlacker

>>kcorbi+(OP)
The major difference between the 2 is how they're being adopted by customers and the tangible value they return.

AI/ML barrier to entry is far simpler and vastly user friendly compared to crypto. Instant value return or gratification from ML products (GTPs and rest) is far more mainstream friendly.

Another view is the "loss" factor. Nobody, thus far, has has had their funds stolen or lost using ML products. I understand content creators and those who, unwillingly, contributed knowledge to learning systems did get circumvented but i'm talking about users/customers. Compare that to the negative stigma of crypto frauds and stereotypical association to illegal transactions.

Apples vs. rotten oranges in my opinion!

>>bamazi+Ye
+1. I have security cameras at home, and my "DVR" is a collection of shell scripts plus a Python script that uses YOLO to find the interesting parts of the footage. The thing works, helps a lot to review daily footage, was damn easy to put to work, and didn't cost me a cent (not even hardware; I run it in a Mini-PC w/o GPU). I knew nothing about ML before writing this script. So yeah, the value is there.

>>epx+gp
Funny, I built a security camera setup with PoE CCTV cameras, GStreamer and the NVIDIA CUDA element using a Xavier platform. I tried SSDMobileNet and YOLO and found them to be absolutely horrible.

The camera that was pointing down at an angle was the worst. Both models would only identify a dog and a person correctly about 15% of the time (missing me or my partner as I walked by and waved), with an actual object detection about 80% of the time even when there was nothing in its ground truths in-frame!! (usually as desks, beds or chairs, i don't recall exactly but it was furniture - and it was pointed at my empty back lot). It had just as many shadow/sunspot/tree failures as Motion. The other camera at eye level did a great job with cars, but not so much with people's side profiles, only head-on.

It was laughably bad. And I have no intention of training my own models on my datasets because I don't have time to label. I did this in 2018-2019 so I don't know what the state of the art object detection models are like today, maybe they got their shit together for non-canonical angles.

I eventually switched back to full-time recording on a 2 TB HDD and if I need to scan back i can jog the livestream because it saves weeks of data.