Mask-Net: Learning Context Aware Invariant Features Using Adversarial Forgetting (Student Abstract)

AAAI 2023 pp. 16374-16375

doi:10.1609/AAAI.V37I13.27047 /aaai/2023/yadav2023aaai-mask/

Abstract

Training a robust system, e.g., Speech to Text (STT), requires large datasets. Variability present in the dataset, such as unwanted nuances and biases, is the reason for the need for large datasets to learn general representations. In this work, we propose a novel approach to induce invariance using adversarial forgetting (AF). Our initial experiments on learning invariant features such as accent on the STT task achieve better generalizations in terms of word error rate (WER) compared to traditional models. We observe an absolute improvement of 2.2% and 1.3% on out-of-distribution and in-distribution test sets, respectively.

PDF AAAI Semantic Scholar

Cite

Text

Yadav and Shah. "Mask-Net: Learning Context Aware Invariant Features Using Adversarial Forgetting (Student Abstract)." AAAI Conference on Artificial Intelligence, 2023. doi:10.1609/AAAI.V37I13.27047

Markdown

[Yadav and Shah. "Mask-Net: Learning Context Aware Invariant Features Using Adversarial Forgetting (Student Abstract)." AAAI Conference on Artificial Intelligence, 2023.](https://mlanthology.org/aaai/2023/yadav2023aaai-mask/) doi:10.1609/AAAI.V37I13.27047

BibTeX

@inproceedings{yadav2023aaai-mask,
  title     = {{Mask-Net: Learning Context Aware Invariant Features Using Adversarial Forgetting (Student Abstract)}},
  author    = {Yadav, Hemant and Shah, Rajiv Ratn},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2023},
  pages     = {16374-16375},
  doi       = {10.1609/AAAI.V37I13.27047},
  url       = {https://mlanthology.org/aaai/2023/yadav2023aaai-mask/}
}