Deep spoken keyword spotting: An overview
Spoken keyword spotting (KWS) deals with the identification of keywords in audio streams
and has become a fast-growing technology thanks to the paradigm shift introduced by deep …
and has become a fast-growing technology thanks to the paradigm shift introduced by deep …
[PDF][PDF] Semi-orthogonal low-rank matrix factorization for deep neural networks.
Abstract Time Delay Neural Networks (TDNNs), also known as onedimensional
Convolutional Neural Networks (1-d CNNs), are an efficient and well-performing neural …
Convolutional Neural Networks (1-d CNNs), are an efficient and well-performing neural …
Hello edge: Keyword spotting on microcontrollers
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 …
on smart devices. It requires real-time response and high accuracy for good user …
Survey on machine learning in speech emotion recognition and vision systems using a recurrent neural network (RNN)
This is a survey paper that aims to give reviews about that finest architectures of machine
learning, the use of algorithms and the applications of the system and speech and vision …
learning, the use of algorithms and the applications of the system and speech and vision …
Survey on deep neural networks in speech and vision systems
This survey presents a review of state-of-the-art deep neural network architectures,
algorithms, and systems in speech and vision applications. Recent advances in deep …
algorithms, and systems in speech and vision applications. Recent advances in deep …
Temporal convolution for real-time keyword spotting on mobile devices
Keyword spotting (KWS) plays a critical role in enabling speech-based user interactions on
smart devices. Recent developments in the field of deep learning have led to wide adoption …
smart devices. Recent developments in the field of deep learning have led to wide adoption …
Convolutional recurrent neural networks for small-footprint keyword spotting
Keyword spotting (KWS) constitutes a major component of human-technology interfaces.
Maximizing the detection accuracy at a low false alarm (FA) rate, while minimizing the …
Maximizing the detection accuracy at a low false alarm (FA) rate, while minimizing the …
[HTML][HTML] Multi-task learning and weighted cross-entropy for DNN-based keyword spotting
Abstract We propose improved Deep Neural Network (DNN) training loss functions for more
accurate single keyword spotting on resource-constrained embedded devices. The loss …
accurate single keyword spotting on resource-constrained embedded devices. The loss …
Grape bunch detection at different growth stages using deep learning quantized models
The agricultural sector plays a fundamental role in our society, where it is increasingly
important to automate processes, which can generate beneficial impacts in the productivity …
important to automate processes, which can generate beneficial impacts in the productivity …
[HTML][HTML] Compressed time delay neural network for small-footprint keyword spotting
In this paper we investigate a time delay neural network (TDNN) for a keyword spotting task
that requires low CPU, memory and latency. The TDNN is trained with transfer learning and …
that requires low CPU, memory and latency. The TDNN is trained with transfer learning and …