Talk:Private biometrics

This article seems promotional, especially when it claims to have created a homomorphic encryption. Their website contains reference to a paper they wrote, and claims they created a one-way encryption which supports a measure of distance. Even at the level of basic definitions, this would not qualify as a fully homomorphic cryptosystem. Rather it is much more like this more modest article: a method to measure distance between two faces or voices.213.8.204.44 (talk) 17:03, 11 January 2019 (UTC)[reply]

That's a kind way to put it. The entire article seems to exist to establish a relationship between the above mentioned paper/Private.id and cryptography, by mentioning encryption systems, homomorphic encryption, and naming cryptographic hash functions. However, not by any stretch of the imagination is the main "solution" put forward by the article related to cryptography, which is summarized in the following sentence:

"A promising method of homomorphic encryption on biometric data is the use of machine learning models to generate feature vectors. For black-box models, such as neural networks, these vectors can not by themselves be used to recreate the initial input data and are therefore a form of one-way encryption."

There is no literature to support that a machine learning model would be equivalent to a homomorphic cryptosystem. Quitlox (talk) 13:10, 26 September 2023 (UTC)[reply]

Eu[edit]

Blb 2A02:2F01:7D13:8E00:D163:C28F:56A7:F29D (talk) 13:42, 3 May 2022 (UTC)[reply]