Deep_Learning_for_Face_Recognition.pdf

Transcript

1. Face Recognition on mobile applications Deep Learning & Augmented Reality
2. What is Deep Learning ? What is Deep Learning ? Deep Learning is a subfield of Machine Learning : A specific way of learning representations from data that puts an emphasis on learning successive layers of increasingly meaningful representations Other approaches to Machine Learning tend to focus on learning only one or two layers of representations of the data -> Shallow Learning Deep Learning = Deep Sequence of simple data transformations (layers) Layered representations of the data are almost always learned via models called Neural Networks
3. Convolutional Neural Networks for image Recognition Source : https://www.mathworks.com/discovery/convolutional-neural-network.html
4. Why do we need convnets ? Convolutional Neural Network (convnet) ➢ A densely-connected neural network has to learn the same “cat features” in different areas ◦ What if cats appear in a different place in the test set ? Source : https://www.coursera.org/learn/intro-to-deep-learning
5. Defining the convolutional operation Convolutional Neural Network (convnet) Source : https://www.coursera.org/learn/intro-to-deep-learning Global vs Local patterns ➢ A Dense layer learns global patterns (involving all pixels) ➢ A Convolutional layer learns local patterns
6. Key characteristics of convolutional neural networks Convolutional Neural Network (convnet) ➢ The patterns they learn are translation invariant: it can recognize a pattern anywhere ➢ They can learn spatial hierarchies of patterns ◦ A first convolutional layer will learn small local patterns such as edges ◦ A second convolutional layer will learn larger patterns made of the features of the first layers, and so on. ◦ Allows convnets to learn increasingly complex and abstract visual concepts
7. What do we need to train a convnet for object recognition ? Training a Convolutional Neural Network Lots of data A CNN architecture A powerful machine
8. Deep Learning for Face Recognition
9. Convnets for Face Recognition Deep Learning for Face Recognition Applying an object recognition architecture ➢ Each person is a class ➢ Need for a lot of photos for each person in order to be able to distinguish them Problems it raises ➢ Usually don’t have more than 1 or 2 photos of the persons we want to recognize ➢ If we want to recognize a new person, we would have to re-train the whole network Does it work the same ?
10. Deep Learning for Face Recognition
11. Using a convnet to compute image embeddings Deep Learning for Face Recognition Vector representation of the image
12. Siamese Networks Deep Learning for Face Recognition Similarity score
13. Training a Convolutional Neural Network for Face Recognition
14. Celebrities Face Dataset What data to train on ? ➢ Training : Need for a few photos of some persons ➢ Inference : Only one photo of the targeted persons to compute the features vector which will be the reference representation to compare with new photos
15. Deep Learning for Face Recognition
16. References References ▼ https://github.com/nyoki-mtl/keras-facene t REF