Hacker News Reader

Comment by dannyw | Hacker News Reader

And you think this is an appropriate use of facial recognition?

Actually I think it's a perfectly appropriate use of facial recognition in this case.

Firstly, the school district already has a picture of every child at the school. Second, they already track the child (manually) at all times throughout the day.

So, they are not collecting any more information than they already had, nor are they collecting any more information than is legally required (attendance tracking is likely to be legally mandated)

The end result is the same exact data exposure / privacy risk that they started with, and more classroom hours to actually teach instead of taking attendance.

The classroom setting is a particularly good example of the narrow use case when facial recognition is acceptable. Unlike, for example, in a retail establishment, where a casual customer who browses the store but does not buy anything would otherwise remain anonymous.

Quite the contrary, you do not want anyone in the classroom who is anonymous.

Speaking purely on the technical side of this problem: if you had to design a system to take attendance without some sort of human in the loop (e.g. a teacher), what other workable solutions are there? ID scanned by itself? A student could scan multiple IDs for their friends. ID + PIN? Due to the bursty nature of entering classrooms, that would cause a jam up at the PIN reader. Short of an ID shackle (that's a joke), nothing really comes to my mind that is foolproof.

I've worked as a teacher before. I'm a bit confused as to the 17,000 hours time claimed to collect attendance information (and hence the justification that you need to automate it). Where is that number coming from?

As part of my job, I had to be able to recognise and put a name to all of my students. This is legitimately difficult and I spent a lot of time doing it. However, it was not for doing attendance! If you are in a school system below university and your teacher doesn't recognise you or know your name... I mean... I can't even comprehend what a bad situation that is. How could you ever teach a student if you know so little about them that you can't even recognise them? When you are marking their paper and see their name, if you don't even know who they are, how can you give appropriate guidance? That kind of situation would be shocking. And to be fair... because it is really, really hard I know a lot of teachers who have no clue... but seriously, this is not what we aim for!

Once you know the students, attendance is easy. Why is that chair empty? Who is supposed to sit there? Look at your seating chart (seriously, it's the best way to learn their names...). Is that person anywhere else in the room? No? They are absent. Mark it on your list. That takes all of 2 seconds. If you are really doing your job, you can ask why the person is missing. Maybe you can have their friend take them some homework. A surveillance system can't do that.

I mean... Why would you ever want an automated system in this circumstance? It just makes no sense.

I think the reality of the situation is that they really wanted a surveillance system. They didn't want to record attendance, they wanted to know where the students were in every second of the day. They wanted to time home long they are in the toilet so that they can catch them smoking or dealing drugs. They wanted to see them leaving the school o that they can lie in wait and nab them as they cut class: students 2 and 5 are supposed to be in band class but they are headed in the opposite direction -- go all you zigs! For great justice!

I just can't imagine any other reason you'd want this. Perhaps I'm wrong. I certainly hope so!

The 17,000 number does seem incomprehensibly high unless it's across the whole district. I'm not a teacher but I'm guessing the attendance gets compiled together and stored? Maybe this starts from hardcopy and includes all the work until its in the final digitalized format?

>They didn't want to record attendance, they wanted to know where the students were in every second of the day. They wanted to time home long they are in the toilet so that they can catch them smoking or dealing drugs. They wanted to see them leaving the school o that they can lie in wait and nab them as they cut class: students 2 and 5 are supposed to be in band class but they are headed in the opposite direction -- go all you zigs! For great justice!

GDPR explicitly prevents this. Data is collected for a purpose and that purpose only. Collecting data for attendance then using it for catching drug dealers is forbidden.

I suspect this is why they are being fined, but the article is a little bit vague on this unfortunately.

If I had to design a system to track students without teacher's involvement, my first step would be to find the stakeholders and remind them that they are asking for a system that is worse, more expensive, and more dystopian than the simple system they already have. It's kind of a false question - like asking, "suppose that a boot was going to stamp on the face of humanity forever, what kind of leather would you make it out of?"

I'm not a proponent of facial recognition at all and would not endorse it especially in this use case. But, if you believe the 17,000 hours/year number, that's ~$500,000/year (assuming $30/hour, random guess) in potential savings. Are you saying there can't possibly be system that could be designed to take attendance while not further infringing on a student's privacy that's cheaper than that?

What is probably happening is that this is some sneaky way of introducing facial recognition in schools and something benign like attendance was the excuse. That, however, is a separate conversation from my original post of how would you automate attendance without using facial recognition.

When trying to design technical solutions, focusing purely on the technical side of the problem, without considering the other aspects, is what led to pervasive tracking and legislation against it in the first place.

Reply on news.ycombinator.com