I recently had a lot of issues with howdy (bugs, dependencies issues, etc.) and the project seems to be not actively maintained anymore.
Therefore, I did start a new software in rust: eyMate.
The main benefit from eyMate is that it uses facenet instead of dlib and rust instead of python. This results in faster and more reliable detection. Also, eyMate allows the usage of a normal cam instead of an IR sensor for face detection.
The project does work, but there are still some missing features (for example: multiple detection tensors for one user) and most certainly some bugs (even major ones). Another important notice is that I could only test the arch package and not the Debian or Fedora ones, those will need some testing by someone else.
It would be great if someone would take a look and provide feedback (or even a PR). This way the project might get to a fully working stage. For discussions, you also might take a look here.
I recently had a lot of issues with howdy (bugs, dependencies issues, etc.) and the project seems to be not actively maintained anymore.
Therefore, I did start a new software in rust: eyMate.
The main benefit from eyMate is that it uses facenet instead of dlib and rust instead of python. This results in faster and more reliable detection. Also, eyMate allows the usage of a normal cam instead of an IR sensor for face detection.
The project does work, but there are still some missing features (for example: multiple detection tensors for one user) and most certainly some bugs (even major ones). Another important notice is that I could only test the arch package and not the Debian or Fedora ones, those will need some testing by someone else.
It would be great if someone would take a look and provide feedback (or even a PR). This way the project might get to a fully working stage. For discussions, you also might take a look here.