Implicit acoustic echo cancellation for keyword spotting and device-directed speech detection

In many speech-enabled human-machine interactions, user speech can overlap with the device playback audio. In these instances, the performance of tasks such as keyword-spotting (KWS) and device-directed speech detection (DDD) can de- grade significantly. To address this problem, we propose an implicit acoustic echo cancellation (iAEC) framework where a neural network is trained to exploit the additional informa- tion from a reference microphone channel to learn to ignore the interfering signal and improve detection performance. We study this framework for the tasks of KWS and DDD on, respectively, an augmented version of Google Speech Com- mands v2 and a real-world Alexa device dataset. Notably, we show a 56% reduction in false-reject rate for the DDD task during device playback conditions. We also show compara- ble or superior performance over a strong end-to-end neural echo cancellation baseline for the KWS task with two order of magnitude less computational requirements.