Challenge entrants are supplied with a fully functioning baseline system. Figure 1 shows a simplified schematic:
- A scene generator (blue box) creates speech in noise signals for a set of six hearing aid microphone inputs.
- The enhancement stage (pink box) takes multichannel hearing aid inputs and attempts to enhance the target speech source, outputting a left and right channel.
- The hearing aid amplification stage (yellow box) applies listener-specific amplification and compression following a standard hearing aid fitting.
- Listener characteristics. i.e. an audiogram (green oval) are supplied to both the enhancement and amplification stage.
- The hearing aid outputs are then evaluated by computing the mean of the Hearing-Aid Speech Perception Index (HASPI)  and Hearing-Aid Speech Quality Index (HASQI)  (orange boxes) which are estimates of intelligibility and quality respectively . Note, HASPI and HASQI are 'intrusive' measures which require a clean speech reference as indicated in the figure.
Your challenge is to improve what happens in the pink enhancement box. The rest of the baseline is fixed and should not be changed.
More details of the different parts of the baseline appear on the core software page, see,
The code for the baseline system, and all supporting Clarity code, is available on GitHub.
Baseline performance using amplification with no enhancement will appear shortly.
- Kates, J.M. and Arehart, K.H., 2021. The hearing-aid speech perception index (HASPI) version 2. Speech Communication, 131, pp.35-46.
- Kates, J.M. and Arehart, K.H., 2014. "The hearing-aid speech quality index (HASQI) version 2". Journal of the Audio Engineering Society. 62 (3): 99–117.