mlsanityData to 7 Oct 2026

Query-by-Example Keyword Spotting Using Spectral-Temporal Graph Attentive Pooling and Multi-Task Learning

arXiv 2409.00099 ↗First posted 27 Aug 2024cs.CL cs.AI cs.SDcited by 0

Not on the map. Why?

Abstract (arXiv)

Existing keyword spotting (KWS) systems primarily rely on predefined keyword phrases. However, the ability to recognize customized keywords is crucial for tailoring interactions with intelligent devices. In this paper, we present a novel Query-by-Example (QbyE) KWS system that employs spectral-temporal graph attentive pooling and multi-task learning. This framework aims to effectively learn speaker-invariant and linguistic-informative embeddings for QbyE KWS tasks. Within this framework, we investigate three distinct network architectures for encoder modeling: LiCoNet, Conformer and ECAPATDNN. The experimental results on a substantial internal dataset of 629 speakers have demonstrated the effectiveness of the proposed QbyE framework in maximizing the potential of simpler models such as LiCoNet. Particularly, LiCoNet, which is 13x more efficient, achieves comparable performance to the computationally intensive Conformer model (1.98% vs. 1.63% FRR at 0.3 FAs/Hr).

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 19 of its 29 references that are in the map. First posted on arXiv 27 Aug 2024.

Builds on

19 of 29 references.

References · 19

  1. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,773
  2. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  3. 2015Cyclical Learning Rates for Training Neural Networkscited by 627
  4. 2020ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verificationcited by 606
  5. 2019Attention Augmented Convolutional Networkscited by 363
  6. 2020Transformer Transducer: A Streamable Speech Recognition Model with Transformer Encoders and RNN-T Losscited by 232
  7. 2019SoftTriple Loss: Deep Metric Learning Without Triplet Samplingcited by 163
  8. 2020Emformer: Efficient Memory Transformer Based Acoustic Model For Low Latency Streaming Speech Recognitioncited by 111
  9. 2018Attentive Statistics Pooling for Deep Speaker Embeddingcited by 108
  10. 2019QuartzNet: Deep Automatic Speech Recognition with 1D Time-Channel Separable Convolutionscited by 93
  11. 2019Jasper: An End-to-End Convolutional Neural Acoustic Modelcited by 75
  12. 2021End-to-End Spectro-Temporal Graph Attention Networks for Speaker Verification Anti-Spoofing and Speech Deepfake Detectioncited by 68
7 more
  1. 2019Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognitioncited by 26
  2. 2021Query-by-Example Keyword Spotting system using Multi-head Attention and Softtriple Losscited by 25
  3. 2018Attention-based End-to-End Models for Small-Footprint Keyword Spottingcited by 18
  4. 2021Graph Attention Networks for Anti-Spoofingcited by 17
  5. 2020Small-Footprint Open-Vocabulary Keyword Spotting with Quantized LSTM Networkscited by 9
  6. 2022QbyE-MLPMixer: Query-by-Example Open-Vocabulary Keyword Spotting using MLPMixercited by 4
  7. 2022LiCo-Net: Linearized Convolution Network for Hardware-efficient Keyword Spottingcited by 4

Built on by

Not cited yet.