Language Identification Using Deep Convolutional Recurrent Neural Networks
Language Identification (LID) systems are used to classify the spoken language from a given audio sample and are typically the first step for many spoken language processing tasks, such as Automatic Speech Recognition (ASR) systems. Without automatic language detection, speech utterances cannot be parsed correctly and grammar rules cannot be applied, causing subsequent speech recognition steps to fail. We propose a LID system that solves the problem in the image domain, rather than the audio domain. We use a hybrid Convolutional Recurrent Neural Network (CRNN) that operates on spectrogram images of the provided audio snippets. In extensive experiments we show, that our model is applicable to a range of noisy scenarios and can easily be extended to previously unknown languages, while maintaining its classification accuracy. We release our code and a large scale training set for LID systems to the community.
PDF AbstractCode
Datasets
Results from the Paper
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Spoken language identification | YouTube News dataset (Background Music) | Inception-v3 CRNN | F1 Score | 0.89 | # 1 | |
Accuracy | 0.89 | # 1 | ||||
Spoken language identification | YouTube News dataset (Background Music) | CRNN | F1 Score | 0.70 | # 2 | |
Accuracy | 0.70 | # 2 | ||||
Spoken language identification | YouTube News dataset (Crackling Noise) | Inception-v3 CRNN | F1 Score | 0.93 | # 1 | |
Accuracy | 0.93 | # 1 | ||||
Spoken language identification | YouTube News dataset (Crackling Noise) | CRNN | F1 Score | 0.83 | # 2 | |
Accuracy | 0.82 | # 2 | ||||
Spoken language identification | YouTube News dataset (No Noise) | CRNN | F1 Score | 0.91 | # 2 | |
Accuracy | 0.91 | # 5 | ||||
Spoken language identification | YouTube News dataset (No Noise) | Inception-v3 CRNN | F1 Score | 0.96 | # 1 | |
Accuracy | 0.96 | # 3 | ||||
Spoken language identification | YouTube News dataset (White Noise) | Inception-v3 CRNN | F1 Score | 0.91 | # 1 | |
Accuracy | 0.91 | # 2 | ||||
Spoken language identification | YouTube News dataset (White Noise) | CRNN | F1 Score | 0.63 | # 2 | |
Accuracy | 0.63 | # 5 |