A. design neural networks
- Going deeper with convolutions
- Rethinking the Inception Architecture for Computer Vision
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning
- Deep Residual Learning for Image Recognition
- Residual Networks are Exponential Ensembles of Relatively Shallow Networks
- Speed/accuracy trade-offs for modern convolutional object detectors
- Dropout: A simple way to prevent neural networks from overfitting
B. Autoencoders
- Auto-Encoding Variational Bayes
- Tutorial on Variational Autoencoders
- Autoencoders, Unsupervised Learning, and Deep Architectures
- A practical tutorial on autoencoders for nonlinear feature fusion: Taxonomy, models, software and guidelines
- Autoencoding beyond pixels using a learned similarity metric
- DeepCoder: Semi-parametric Variational Autoencoders for Automatic Facial Action Coding
- Extracting and Composing Robust Features with Denoising Autoencoders
C. Neural Networks compression
- Distilling the Knowledge in a Neural Network
- Model Distillation with Knowledge Transfer from Face Classification to Alignment and Verification
- Fitnets: hints for thin deep nets
- Born-Again Neural Network
- Deep Model Compression: Distilling Knowledge from Noisy Teachers
- Do Deep Convolutional nets really need to be deep and convolutional?
- Compression of deep convolutional neural networks for fast and low power mobile appplications
D. Face recognition
- OpenFace: A general-purpose facce recognition library with mobile applications
- FaceNet: A Unified Embedding for Face Recognition and Clustering
- Deep metric learning using triplet network
E. Super Resolution
- Super-Resolution from a single image
- Real-Time Single image and video super-resolution using an efficient sub-pixel convolutional neural network
- image restoration using convolutional auto-encoders with symmetric skip connections
F. preprocessomg
- One millisecond face alignment with an Ensemble of regression trees
G.