Spoken Language Identification using ConvNets
This work addresses language identification for speech processing systems like voice assistants, but it is incremental as it builds on existing models with a new attention mechanism and feature exploration.
The paper tackled the problem of spoken language identification without transcriptive data by benchmarking existing models and proposing a new attention-based model using log-Mel spectrogram images and raw waveforms as features, achieving accuracies of 95.4% for six languages and 96.3% for four languages on the VoxForge dataset.
Language Identification (LI) is an important first step in several speech processing systems. With a growing number of voice-based assistants, speech LI has emerged as a widely researched field. To approach the problem of identifying languages, we can either adopt an implicit approach where only the speech for a language is present or an explicit one where text is available with its corresponding transcript. This paper focuses on an implicit approach due to the absence of transcriptive data. This paper benchmarks existing models and proposes a new attention based model for language identification which uses log-Mel spectrogram images as input. We also present the effectiveness of raw waveforms as features to neural network models for LI tasks. For training and evaluation of models, we classified six languages (English, French, German, Spanish, Russian and Italian) with an accuracy of 95.4% and four languages (English, French, German, Spanish) with an accuracy of 96.3% obtained from the VoxForge dataset. This approach can further be scaled to incorporate more languages.