The paper presents a method for spoken term detection based on the
Transformer architecture. We propose the encoder-encoder architecture employing
two BERT-like encoders with additional modifications, including convolutional
and upsampling layers, attention masking, and shared parameters. The encoders
project a recognized hypothesis and a searched term into a shared embedding
space, where the score of the putative hit is computed using the calibrated dot
product. In the experiments, we used the Wav2Vec 2.0 speech recognizer, and the
proposed system outperformed a baseline method based on deep LSTMs on the
English and Czech STD datasets based on USC Shoah Foundation Visual History
Archive (MALACH).

本文提出了一种基于 Transformer 结构的语音词项检测方法，采用两个 BERT-like 编码器并进行了适当的修改，包括卷积和上采样层、注意力掩码、共享参数等，将假设和搜索术语投影到共享嵌入空间中，并使用校准点积计算假设的命中得分。在实验中，使用了 Wav2Vec 2.0 语音识别器，并在基于 USC Shoah 基金会视觉历史档案（MALACH）的英语和捷克 STD 数据集上，提出的系统优于基于深度 LSTM 的基线方法。