musa-atlihan/autonomous-vehicle-path

Reading list for autonomous driving technology.

Reading List for Autonomous Driving

Deep Learning and Computer Vision

Transfer Learning

• M. D. Zeiler, R. Fergus, "Visualizing and Understanding Convolutional Networks", arXiv, 2013.

  • Visualizations for convolutional layers in order to understand the hierarchy among layers. This article helps understand why transfer learning is applicable for convolutional neural networks.

• J. Yosinski, J. Clune, Y. Bengio, H. Lipson, "How transferable are features in deep neural networks?", arXiv, 2014.

  • A very extensive investigation on transfer learning. ImageNet dataset is split into two small datasets (by randomly chosing the classes and in addition, chosing the clases based on human made and nature categories). The general and specific features are successfully well defined by using one of the datasets as the control. Effects of fine-tuning the different layers are discussed.