2 ms·
The resulting image would be basically useless for navigation. If a human can't make stuff out in the image, a computer definitely won't.
by stavros 2y ago
The resulting image would be basically useless for navigation. If a human can't make stuff out in the image, a computer definitely won't.
- strken 2y agoWhy would that be true? You can do facial recognition on very low resolution photographs that wouldn't be easily recognisable to a human: https://www.sciencedirect.com/science/article/abs/pii/S0262885619300216 https://www.sciencedirect.com/science/article/abs/pii/S02628...
- gwbas1c 2y agoThat can be achieved by manipulating the lens to blur the image. A blurry high resolution image is generally equivalent to a low resolution image.
- pbhjpbhj 2y agoI was thinking more of an equivalent of XOR-ing the image direct from the CCD then using a TPM to do the image processing (edge detection, or whatever). You could deobfuscate by inspecting an individual CCD, but all images passed around would be essentially white-noise?
- stavros 2y agoYou trust the company to do all this complicated obfuscation, but don't trust them to not spy on you?
- textlapse 2y agoAll of the HW components are commoditized and horizontally integrated anyways. It’s easy to verify and build components with strong guarantees that ensures that the integrator can’t workaround (and why would they?) I do worry that companies essentially use the webcam as their main profit margin (sell ads!) with a rubbish trash compactor added as a mere ‘free service’.