Publications Repository - Gdańsk University of Technology

This paper describes two novel complementary techniques that improve the detection of lexical stress errors in non-native (L2) English speech: attention-based feature extraction and data augmentation based on Neural Text-To-Speech (TTS). In a classical approach, audio features are usually extracted from fixed regions of speech such as the syllable nucleus. We propose an attention-based deep learning model that automatically de rives optimal syllable-level representation from frame-level and phoneme-level audio features. Training this model is challenging because of the limited amount of incorrect stress patterns. To solve this problem, we propose to augment the training set with incorrectly stressed words generated with Neural TTS. Combining both techniques achieves 94.8% precision and 49.2% recall for the detection of incorrectly stressed words in L2 English speech of Slavic and Baltic speakers.

Authors

• Daniel Korzekwa link open in new tab ,
• prof. Roberto Barra-Chicote,
• mgr inż. Szymon Zaporowski link open in new tab ,
• Grzegorz Beringer,
• Jaime Lorenzo-trueba,
• Alicja Serafinowicz,
• Jasha Droppo,
• dr Thomas Drugman,
• prof. dr hab. inż. Bożena Kostek link open in new tab