Facial Recognition
Hi, I'm about to start my final year of sixth form (British equivalent of a senior in high school) and was wondering if I could get some help with sorting out the logistics of my programme.
I essentially need a low resource intensive method of training a partially pre-trained model with part of my own dataset of faces.
My plan is to create the UI in C# but handle the facial recognition in python.
I was wondering if any of you have any idea where I can get resources to sort of walk me through - only issue is I can not use a library that handles everything for me as it takes the complexity out of my programme which is required for a passing grade.
The programme is essentially :
Users can log-in or sign-up - if they're signing up, the programme will ask them to take 7 images ( to cover 180 degrees) and they'll be able to seem themselves through live webcam footage. They'll have to press capture each time and there will be a 5 second timer for them to get ready after each photo. If they are already signed up, they can log-in by pressing log-in and then a capture button.
They should also have a sign-out option.
A sign in and sign out will be result to data being added to a txt file or excel spreadsheet with a timestamp and a persons name
The real complexity comes from using a CNN and training it to recognise faces (I'm allowed to use a library for this) and then training it to recognise particular individuals. I don't have a GPU with my laptop and my school has some gaming pcs but I wouldn't be able to use them over the summer (My school closes next week) so I would need a non-resource intensive method
Any advice would be appreciated and feel free to ask question to solve any confusion
1 Reply
So far I've only managed to find this : https://www.width.ai/post/tensorflow-facial-recognition
How to Implement TensorFlow Facial Recognition From Scratch | Width.ai
Get a simple TensorFlow facial recognition model up & running quickly with this tutorial aimed at using it in your personal spaces on smartphones & IoT devices.