Breaking Down Power Barriers in On-Device Streaming ASR: Insights and Solutions
This work addresses power efficiency for on-device speech recognition users, offering a domain-specific incremental improvement.
The study tackled the problem of high power consumption in on-device streaming speech recognition by analyzing how weight parameters affect energy efficiency, and proposed design principles that reduce energy usage by up to 47% while maintaining accuracy and improving real-time performance.
Power consumption plays a crucial role in on-device streaming speech recognition, significantly influencing the user experience. This study explores how the configuration of weight parameters in speech recognition models affects their overall energy efficiency. We found that the influence of these parameters on power consumption varies depending on factors such as invocation frequency and memory allocation. Leveraging these insights, we propose design principles that enhance on-device speech recognition models by reducing power consumption with minimal impact on accuracy. Our approach, which adjusts model components based on their specific energy sensitivities, achieves up to 47% lower energy usage while preserving comparable model accuracy and improving real-time performance compared to leading methods.