A study on more realistic room simulation for far-field keyword spotting
This work addresses the challenge of robust keyword spotting in noisy far-field environments for speech recognition systems, representing an incremental improvement in simulation techniques.
The study tackled the problem of improving far-field keyword spotting by using more realistic room simulation for training, resulting in up to 35.8% relative improvement over a common baseline method without fine-tuning on in-domain data.
We investigate the impact of more realistic room simulation for training far-field keyword spotting systems without fine-tuning on in-domain data. To this end, we study the impact of incorporating the following factors in the room impulse response (RIR) generation: air absorption, surface- and frequency-dependent coefficients of real materials, and stochastic ray tracing. Through an ablation study, a wake word task is used to measure the impact of these factors in comparison with a ground-truth set of measured RIRs. On a hold-out set of re-recordings under clean and noisy far-field conditions, we demonstrate up to $35.8\%$ relative improvement over the commonly-used (single absorption coefficient) image source method. Source code is made available in the Pyroomacoustics package, allowing others to incorporate these techniques in their work.