Deep Learning with Pytorch in a Nutshell
  • Content
  • Image recognition
  • Object detection
  • Semantic segmentation
  • GAN
  • Image style transfer
  • Face recognition
  • Interpretability
  • Word embedding
  • Pytorch
  • Optimization
  • Special layers
    • Transposed convolution
  • Neural architecture search
  • Reinforcement learning
    • Proof of Bellman equation
    • Tabular solution method
Powered by GitBook
On this page
  • Learning a Similarity Metric Discriminatively, with Application to Face Verification (Siamese network)
  • DeepFace: Closing the Gap to Human-Level Performance in Face Verification
  • FaceNet: A Unified Embedding for Face Recognition and Clustering

Face recognition

PreviousImage style transferNextInterpretability

Last updated 6 years ago

Learning a Similarity Metric Discriminatively, with Application to Face Verification (Siamese network)

Training a similarity metric from data

DeepFace: Closing the Gap to Human-Level Performance in Face Verification

FaceNet: A Unified Embedding for Face Recognition and Clustering

∣∣f(A(i))−f(P(i))∣∣22+α<∣∣f(A(i))−f(N(i))∣∣22||f(A^{(i)})-f(P^{(i)})||_2^2 + \alpha \lt ||f(A^{(i)})-f(N^{(i)})||_2^2∣∣f(A(i))−f(P(i))∣∣22​+α<∣∣f(A(i))−f(N(i))∣∣22​
Triplet loss=[∣∣f(A(i))−f(P(i))∣∣22−∣∣f(A(i))−f(N(i))∣∣22+α]+\text{Triplet loss}= \Big[||f(A^{(i)})-f(P^{(i)})||_2^2 - ||f(A^{(i)})-f(N^{(i)})||_2^2 + \alpha \Big]_+Triplet loss=[∣∣f(A(i))−f(P(i))∣∣22​−∣∣f(A(i))−f(N(i))∣∣22​+α]+​

A(i)A^{(i)}A(i) : anchor of i-th triplet, P(i)P^{(i)}P(i): positive example of i-th triplet, N(i)N^{(i)}N(i): negative example of i-th triplet, α\alphaα: margin

http://yann.lecun.com/exdb/publis/pdf/chopra-05.pdf
https://research.fb.com/wp-content/uploads/2016/11/deepface-closing-the-gap-to-human-level-performance-in-face-verification.pdf
https://arxiv.org/abs/1503.03832