3 ms·

But do they have to pair the sensors in each phone distinctly? I assume they have hardware security measures that you cannot just read out the keys from one of

But do they have to pair the sensors in each phone distinctly? I assume they have hardware security measures that you cannot just read out the keys from one of the sensors. So why not just put the same "key" in all of the official IR cameras? Then you could still swap cameras with official safe ones, but you could not build a fake replacement.

neilalexander 5y ago
The sensors are using unique patterns on each device. There would need to be a pairing between the TrueDepth camera and the Secure Enclave in order to account for this unique pattern. This is why you can't copy Face ID data from one device to another and expect it to work. From the Apple Platform Security document: > After the TrueDepth camera confirms the presence of an attentive face, it projects and reads thousands of infrared dots to form a depth map of the face along with a 2D infrared image. This data is used to create a sequence of 2D images and depth maps which are digitally signed and sent to the Secure Enclave. To counter both digital and physical spoofs, the TrueDepth camera randomises the sequence of 2D images and depth map captures and projects a device-specific random pattern. A portion of the Secure Neural Engine — protected within the Secure Enclave — transforms this data into a mathematical representation and compares that representation with the enrolled facial data. This enrolled facial data is itself a mathematical representation of the user’s face captured across a variety of poses. https://support.apple.com/en-gb/guide/security/sec067eb0c9e/1/web/1 https://support.apple.com/en-gb/guide/security/sec067eb0c9e/...
- captainmuon 5y ago
  So the individual dot patterns are fixed in hardware? That's clever. (A former colleague of mine has a vaguely related patent. They create unique patterns inside of chips by applying (sputtering) material at an angle, which creates chaotic patterns that cannot be copied by lithography, so the chip is in principle unclonable.)

neilalexander 5y ago

The sensors are using unique patterns on each device. There would need to be a pairing between the TrueDepth camera and the Secure Enclave in order to account for this unique pattern. This is why you can't copy Face ID data from one device to another and expect it to work. From the Apple Platform Security document: > After the TrueDepth camera confirms the presence of an attentive face, it projects and reads thousands of infrared dots to form a depth map of the face along with a 2D infrared image. This data is used to create a sequence of 2D images and depth maps which are digitally signed and sent to the Secure Enclave. To counter both digital and physical spoofs, the TrueDepth camera randomises the sequence of 2D images and depth map captures and projects a device-specific random pattern. A portion of the Secure Neural Engine — protected within the Secure Enclave — transforms this data into a mathematical representation and compares that representation with the enrolled facial data. This enrolled facial data is itself a mathematical representation of the user’s face captured across a variety of poses. https://support.apple.com/en-gb/guide/security/sec067eb0c9e/1/web/1 https://support.apple.com/en-gb/guide/security/sec067eb0c9e/...

captainmuon 5y ago
So the individual dot patterns are fixed in hardware? That's clever. (A former colleague of mine has a vaguely related patent. They create unique patterns inside of chips by applying (sputtering) material at an angle, which creates chaotic patterns that cannot be copied by lithography, so the chip is in principle unclonable.)