Conclusion

EfficientMic framework, a novel approach for real-time aliasing detection in low-power embedded systems. By dynamically adjusting the microphone’s sampling rate based on real-time analysis, EfficientMic balances both energy consumption and audio quality, addressing key challenges in resource-constrained settings such as long-lived and energy-harvesting sensor systems. Evaluations on the UrbanSound8K and ESC-50 datasets show that EfficientMic significantly reduces storage and energy consumption, with only a minimal F1-Score drop-as low as 0.7%, outperforming traditional models. Since EfficientMic is designed for automated, on-device aliasing detection, we focus on quantitative metrics like accuracy, energy, and latency rather than STFT analysis. In the future, we plan to explore the effect of highly accurate aliasing detection on the quality of STFTs to determine the impact of aliasing suppression. Overall, EfficientMic offers an efficient, low-latency solution for real-time audio processing, making it well-suited for deployment in embedded and power-sensitive environments.