@misc{indiciaeecf6f6505a36,
  title = {Interpreting DNN output layer activations: A strategy to cope with unseen data in speech recognition},
  author = {Vikramjit Mitra and Horacio Franco},
  year = {2018},
  url = {https://arxiv.org/abs/1802.06861},
  note = {Source identifier: 1802.06861}
}