Do Better ImageNet Models Transfer Better? Paper Summary & Analysis

Paper Objective

Background

  1. Fixed feature extraction: the final layer of the Image-Net trained network is removed in favour of a linear classifier, which outputs the class prediction over the classes of the new (target) dataset.
  2. Fine-tuning: the weights of the ImageNet pretrained model are treated as an initialisation for the model trained on the new (target) dataset

Paper Contributions

  1. the absence of a scale parameter (γ) for batch normalization layers
  2. the use of label smoothing
  3. dropout
  4. the presence of an auxiliary classifier head
https://openaccess.thecvf.com/content_CVPR_2019/papers/Kornblith_Do_Better_ImageNet_Models_Transfer_Better_CVPR_2019_paper.pdf

Conclusion

--

--

--

Cornell Data Science is an engineering project team @Cornell that seeks to prepare students for a career in data science.

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Convolutional Neural Models and Multi-Scale Representations

Hands-on Machine Learning in Python — Decision Tree Classification

Cross Validation in Machine Learning

Syncnet Model with VidTIMIT Dataset

How Siri understands what “2+2” is

Amazon data set classification using LR model

The effortless way to process images with OpenCV Canny algorithm.

Understanding Regularisation

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Cornell Data Science

Cornell Data Science

Cornell Data Science is an engineering project team @Cornell that seeks to prepare students for a career in data science.

More from Medium

Reinforcement learning through self-play and populations of agents

Computer vision for in-vitro fertilization (IVF)

How are GPUs used for Deep Learning?