The Sapienza computer scientists say Wi-Fi signals offer superior surveillance potential compared to cameras because they’re not affected by light conditions, can penetrate walls and other obstacles, and they’re more privacy-preserving than visual images.
[…] The Rome-based researchers who proposed WhoFi claim their technique makes accurate matches on the public NTU-Fi dataset up to 95.5 percent of the time when the deep neural network uses the transformer encoding architecture.
ileftreddit@piefed.social 10 hours ago
Why would someone research something like this? God damn, like use your life for good, homie
StenSaksTapir@feddit.dk 5 hours ago
Well I heard about this and thought “this will be great for home automation”, but I also know that someone was equally excited about using this to rob people of basic freedoms or being a fucking creep or both.
gcheliotis@lemmy.world 9 hours ago
I can imagine this being initially an accidental discovery like oh every time so and so’s body interacts with the WiFi signal it’s the same pattern… until someone starts exploring this further… and then some engineer or their manager started looking for applications for this. In my experience engineering researchers especially are very good with coming up with use cases for whatever tech they’re working with, with little ethical consideration.
sugar_in_your_tea@sh.itjust.works 4 hours ago
I doubt it. You’d need to be looking really closely at the waveforms to notice this, so they were likely already doing something similar, like that research that can pinpoint where people are in a house based on their WiFi. They were probably already doing something creepy before they noticed that this was more straightforward than they expected.