mlsanityData to 13 Oct 2026

Hello Edge: Keyword Spotting on Microcontrollers

arXiv 1711.07128 ↗First posted 20 Nov 2017cs.SD cs.CL cs.LGcited by 108

Not on the map. Why?

Abstract (arXiv)

Keyword spotting (KWS) is a critical component for enabling speech based user interactions on smart devices. It requires real-time response and high accuracy for good user experience. Recently, neural networks have become an attractive choice for KWS architecture because of their superior accuracy compared to traditional speech processing algorithms. Due to its always-on nature, KWS application has highly constrained power budget and typically runs on tiny microcontrollers with limited memory and compute capability. The design of neural network architecture for KWS must consider these constraints. In this work, we perform neural network architecture evaluation and exploration for running KWS on resource-constrained microcontrollers. We train various neural network architectures for keyword spotting published in literature to compare their accuracy and memory/compute requirements. We show that it is possible to optimize these neural network architectures to fit within the memory and compute constraints of microcontrollers without sacrificing accuracy. We further explore the depthwise separable convolutional neural network (DS-CNN) and compare it against other neural network architectures. DS-CNN achieves an accuracy of 95.4%, which is ~10% higher than the DNN model with similar number of parameters.

Key facts. Cited by 108 papers in the map (data through 13 Oct 2026). First posted on arXiv 20 Nov 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2021: 22; 2022: 28; 2023: 11; 2024: 17; 2025: 18; 2026: 12.

108 papers.

Sort:

By area:

  1. 2021MLPerf Tiny Benchmarkcited by 122
  2. 2021Keyword Transformer: A Self-Attention Model for Keyword Spottingcited by 60
  3. 2022ZeroFL: Efficient On-Device Training for Federated Learning with Local Sparsitycited by 45
  4. 2021Federated Learning Meets Natural Language Processing: A Surveycited by 43
  5. 2021Deep Spoken Keyword Spotting: An Overviewcited by 43
  6. 2022Machine Learning for Microcontroller-Class Hardware: A Reviewcited by 28
  7. 2021Few-Shot Keyword Spotting in Any Languagecited by 27
  8. 2021Learning Efficient Representations for Keyword Spotting with Triplet Losscited by 24
  9. 2022Asymmetric Temperature Scaling Makes Larger Networks Teach Well Againcited by 23
  10. 2021Quantization and Deployment of Deep Neural Networks on Microcontrollerscited by 21
  11. 2021Widening Access to Applied Machine Learning with TinyMLcited by 17
  12. 2021Differentiable Network Pruning for Microcontrollerscited by 16
38 more
  1. 2024MLPerf Power: Benchmarking the Energy Efficiency of Machine Learning Systems from Microwatts to Megawatts for Sustainable AIcited by 15
  2. 2022Intelligence at the Extreme Edge: A Survey on Reformable TinyMLcited by 13
  3. 2024On-Device Domain Learning for Keyword Spotting on Low-Power Extreme Edge Embedded Systemscited by 13
  4. 2021Visual Keyword Spotting with Attentioncited by 11
  5. 2021Sound Event Detection with Binary Neural Networks on Tightly Power-Constrained IoT Devicescited by 10
  6. 2022Federated Self-supervised Speech Representations: Are We There Yet?cited by 10
  7. 2024Backpropagation through space, time, and the braincited by 10
  8. 2022Channel-wise Mixed-precision Assignment for DNN Inference on Constrained Edge Nodescited by 9
  9. 2022Open-source FPGA-ML codesign for the MLPerf Tiny Benchmarkcited by 9
  10. 2023Few-Shot Open-Set Learning for On-Device Customization of KeyWord Spotting Systemscited by 9
  11. 2023Efficient Neural Networks for Tiny Machine Learning: A Comprehensive Reviewcited by 9
  12. 2022Tiny, always-on and fragile: Bias propagation through design choices in on-device machine learning workflowscited by 8
  13. 2022Rainbow Keywords: Efficient Incremental Learning for Online Spoken Keyword Spottingcited by 8
  14. 2023Enhancing Neural Architecture Search with Multiple Hardware Constraints for Deep Learning Model Deployment on Tiny IoT Devicescited by 7
  15. 2024Self-Learning for Personalized Keyword Spotting on Ultra-Low-Power Audio Sensorscited by 7
  16. 2021Ultra-Low Power Keyword Spotting at the Edgecited by 6
  17. 2023PQA: Exploring the Potential of Product Quantization in DNN Hardware Accelerationcited by 6
  18. 2021Spike-inspired Rank Coding for Fast and Accurate Recurrent Neural Networkscited by 5
  19. 2022Sub-mW Keyword Spotting on an MCU: Analog Binary Feature Extraction and Binary Neural Networkscited by 5
  20. 2022BiFSMN: Binary Neural Network for Keyword Spottingcited by 5
  21. 2022Machine Learning Sensorscited by 5
  22. 2022Adversarial Vulnerability of Randomized Ensemblescited by 5
  23. 2022T-RECX: Tiny-Resource Efficient Convolutional neural networks with early-eXitcited by 5
  24. 2024TinySV: Speaker Verification in TinyML with On-device Learningcited by 5
  25. 2024MATCH: Model-Aware TVM-based Compilation for Heterogeneous Edge Devicescited by 5
  26. 2021Zero-Shot Federated Learning with New Classes for Audio Classificationcited by 4
  27. 2022YONO: Modeling Multiple Heterogeneous Neural Networks on Microcontrollerscited by 4
  28. 2022Depth Pruning with Auxiliary Networks for TinyMLcited by 4
  29. 2022Avoid Overfitting User Specific Information in Federated Keyword Spottingcited by 4
  30. 2022FedorAS: Federated Architecture Search under system heterogeneitycited by 4
  31. 2022Keyword Spotting System and Evaluation of Pruning and Quantization Methods on Low-power Edge Microcontrollerscited by 4
  32. 2022LiCo-Net: Linearized Convolution Network for Hardware-efficient Keyword Spottingcited by 4
  33. 2022BiFSMNv2: Pushing Binary Neural Networks for Keyword Spotting to Real-Network Performancecited by 4
  34. 2023The Case for Hierarchical Deep Learning Inference at the Network Edgecited by 4
  35. 2024DTMM: Deploying TinyML Models on Extremely Weak IoT Devices with Pruningcited by 4
  36. 2024UR2M: Uncertainty and Resource-Aware Event Detection on Microcontrollerscited by 4
  37. 2024Boosting keyword spotting through on-device learnable user speech characteristicscited by 4
  38. 2024QUTE: Quantifying Uncertainty in TinyML with Early-exit-assisted ensembles for model-monitoringcited by 4

50 most cited of 108 All 108 (JSON)