@misc{indiciae53f49d8fe59c,
  title = {Why do linear SVMs trained on HOG features perform so well?},
  author = {Hilton Bristow and Simon Lucey},
  year = {2014},
  url = {https://arxiv.org/abs/1406.2419},
  note = {Source identifier: 1406.2419}
}