A tricked found by Jeremy to avoid overfitting is to train a network with small images for few epochs and then train it using larger images. It is only applicable to architectures that can take arbitrary image sizes and thus not applicable to VGG. (from fast.ai course)
- Explaining CNN by Colah, recommended by Jeremey
- Deep Learning CNN’s in Tensorflow with GPUs
- A technical report on convolution arithmetic in the context of deep learning
- Keras tutorial – build a convolutional neural network in 11 lines
- Image Kernels
- Planet: Understanding the Amazon from Space, 1st Place Winner's Interview
- ResNeXt: Aggregated Residual Transformations for Deep Neural Networks
- Visualizing convolutional neural networks
- An Introduction to different Types of Convolutions in Deep Learning
- Andrew suggests to read papers in the following order: AlexNet -> VGGLeNet
- An Overview of ResNet and its Variants
- Applied Deep Learning - Part 4: Convolutional Neural Networks
- Dilated Convolutions and Kronecker Factored Convolutions
- A Comprehensive Design Guide for Image Classification CNNs
- An intuitive guide to Convolutional Neural Networks
- 3D Visualization of a Convolutional Neural Network
- Intuitively Understanding Convolutions for Deep Learning
- A Simple Guide to the Versions of the Inception Network
- An Interactive Node-Link Visualization of Convolutional Neural Networks
- Understanding Convolutional Neural Networks for NLP – WildML
- Deep Visualization Toolbox - YouTube
- Deep Learning Tips and Tricks – Towards Data Science
- An Intriguing Failing of Convolutional Neural Networks and the CoordConv Solution
- A simple 2D CNN for MNIST digit recognition – Towards Data Science
- How do Convolutional Neural Networks work?
- A Beginner's Guide To Understanding Convolutional Neural Networks – Adit Deshpande – CS Undergrad at UCLA ('19)
- An Intuitive Explanation of Convolutional Neural Networks – the data science blog
- 一文读懂VGG网络
- [1710.05381] A systematic study of the class imbalance problem in convolutional neural networks
- Troubleshooting Convolutional Neural Nets
- Convolutional Neural Networks: The Biologically-Inspired Model
- Troubleshooting Convolutional Neural Nets