Abstract:<p>Keyword search (KWS) means searching for keywords given by the user from continuous speech. Conventional KWS systems are based on Automatic Speech Recognition (ASR), where the input speech has to be first processed by the ASR system before keyword searching. In the recent decade, as deep learning and deep neural networks (DNN) become increasingly popular, KWS systems can also be trained in an end-to-end (E2E) manner. The main advantage of E2E KWS is that there is no need for speech recognition, which makes the training and searching procedure much more straightforward than the traditional ones. This article proposes an E2E KWS model, which consists of four parts: speech encoder-decoder, query encoder-decoder, attention mechanism, and energy scorer. Firstly, the proposed model outperforms the baseline model. Secondly, we find that under various supervision, character or phoneme sequences, speech or query encoders can extract the corresponding information, resulting in different performances. Moreover, we introduce an attention mechanism and invent a novel energy scorer, where the former can help locate keywords. The latter can make final decisions by considering speech embeddings, query embeddings, and attention weights in parallel. We evaluate our model on low resource conditions with about 10-hour training data for four different languages. The experiment results prove that the proposed model can work well on low resource conditions.</p>

Subword scheme for keyword search

Improving keyword search by query expansion in a probabilistic framework

End-to-end keyword search system based on attention mechanism and energy scorer for low resource languages

Keyword-Specific Acoustic Model Pruning for Open-Vocabulary Keyword Spotting

High-Performance Swahili Keyword Search with Very Limited Language Pack: the Thuee System for the Openkws15 Evaluation

U2-KWS: Unified Two-pass Open-vocabulary Keyword Spotting with Keyword Bias

Unrestricted Vocabulary Keyword Spotting Using LSTM-CTC

Linguistic Search Optimization for Deep Learning Based LVCSR

TDT-KWS: Fast And Accurate Keyword Spotting Using Token-and-duration Transducer

Joint Decoding of Tandem and Hybrid Systems for Improved Keyword Spotting on Low Resource Languages

Research on continuous Chinese speech recognition system and knowledge based search strategies

Research on Speaker-Depended Isolated-Word Speech Recognition System

Incorporate Web Search Technology to Solve Out-of-Vocabulary Words in Chinese Word Segmentation.

A Multitask Training Approach to Enhance Whisper with Open-Vocabulary Keyword Spotting

Audio-visual Keyword Spotting for Mandarin Based on Discriminative Local Spatial-Temporal Descriptors.

Sequence discriminative training for deep learning based acoustic keyword spotting

A Dialogue System Based on Keyword Spotting Technology

Subword and Crossword Units for CTC Acoustic Models

A New Approach for Modeling OOV Words.

Web-based keyword adapted Language Modeling for Keyword Spotting

Open vocabulary keyword spotting through transfer learning from speech synthesis