webdizz/upass.ai repository overview

⁠upass.ai

Experiments of IDless person identification with AI

⁠TODO

• Take a look at training dataset VGGFace2⁠
• Pre-process dataset to enable it to be loaded into fastai Learner
• Investigate similar implementations (https://hackernoon.com/facial-similarity-with-siamese-networks-in-pytorch-9642aa9db2f7⁠, https://medium.freecodecamp.org/how-to-train-your-own-faceid-cnn-using-tensorflow-eager-execution-6905afe4fd5a⁠, https://becominghuman.ai/siamese-networks-algorithm-applications-and-pytorch-implementation-4ffa3304c18⁠) to build own one
• Train model
• Collect private dataset