[cs231n Winter2016 Solutions] - CNNs for Visual Recognition

1. Introduction

2. Useful links related

- Syllabus: here
- Lectures: here
- Intro to Neural Networks: here
- Data preprocessing: here
- Parameters update: Vanilla SGD, Momentum, Adam, Adagrad, RMSProp...here
- Xavier/He initialization: here
- L2 reg/ Dropout: here
- Understanding computational graph in ANNs: here
- Detail about analytical gradient in ANNs: here
- Batch-normalization: here
- Back-propagation through Batch-norm using computational graph: here
- Back-propagation through Batch-norm using derivative on paper: here
- ConvNet notes: here
- Understanding RNNs and LSTMs: here
- LeNet-5: here, AlexNet: here, ZFNet: here, VGGNet: here, GoogLeNet: here, ResNet: here

3. Assignments

Here are several exercises we'll going through:
|----[Assignment #1]
     |----knn knn.ipynb
     |----svm svm.ipynb
     |----softmax softmax.ipynb
     |----two_layer_net two_layer_net.ipynb
     |----features features.ipynb
|----[Assignment #2]
     |----FullyConnectedNets FullyConnectedNets.ipynb
     |----BatchNormalization BatchNormalization.ipynb
     |----Dropout Dropout.ipynb
     |----ConvolutionalNetworks ConvolutionalNetworks.ipynb
|----[Assignment #3]
     |----RNN_Captioning RNN_Captioning.ipynb
     |----LSTM_Captioning LSTM_Captioning.ipynb
     |----ImageGradients ImageGradients.ipynb
     |----ImageGeneration ImageGeneration.ipynb

4. Projects

Applied face recognition from the Google's paper: Facenet at here