ASPIRE - Real noisy audio-visual speech enhancement corpus

What is ASPIRE?

ASPIRE is a a first of its kind, audiovisual speech corpus recorded in real noisy environment (such as cafe, restaurants) which can be used to support reliable evaluation of multi-modal Speech Filtering technologies. This dataset follows the same sentence format as the audio-visual Grid corpus. The recorded audiovisual speech corpus can be used for reliable evaluation of next generation multi-modal Speech Filtering technologies.

Citing the corpus

@article{gogate2020cochleanet,
  title={CochleaNet: A robust language-independent audio-visual model for real-time speech enhancement},
  author={Gogate, Mandar and Dashtipour, Kia and Adeel, Ahsan and Hussain, Amir},
  journal={Information Fusion},
  volume={63},
  pages={273--285},
  year={2020},
  publisher={Elsevier}
}

OR

Gogate, M., Dashtipour, K., Adeel, A., & Hussain, A. (2020). CochleaNet: A robust language-independent audio-visual model for real-time speech enhancement. Information Fusion, 63, 273-285.

Acknowledgements

This research was funded by the UK Engineering and Physical Sciences Research Council (EPSRC project AV-COGHEAR, EP/M026981/1)