The repo currently uses a wake word approach to detect whether to put the system in active listening mode.
This approach has multiple issues as wake_word models consume space in memory and are not that accurate.
Therefore the wake_word detention feature must be replaced by a "Hotkey Press" mechanism where whenever a user presses the a defined hotkey the systems comes in active-listening mode.
The repo currently uses a wake word approach to detect whether to put the system in active listening mode.
This approach has multiple issues as wake_word models consume space in memory and are not that accurate.
Therefore the wake_word detention feature must be replaced by a "Hotkey Press" mechanism where whenever a user presses the a defined hotkey the systems comes in active-listening mode.