Table 1 Comparison of AUSpeech with other different databases in terms of resolution, duration, tasks, etc.
From: An Audio-Ultrasound Synchronized Database of Tongue Movement for Mandarin speech
Datasets | No. of speakers | Type | Resolution | Modalities | Language | Duration | Task |
|---|---|---|---|---|---|---|---|
TAL | 82 | Normal | 64 × 842 | Ultrasound, Lip Videos, Audio | English | 13.5 hours(audio) | Sentence, Non-words |
UltraSuite | 86 | Normal/Dysarthria (Children) | 63 × 412 | Ultrasound, Audio, Text | English | 18.67 hours(audio) | Words, Sentence, Non-words |
SSR7000 | 1 | Normal | 640 × 445 | Ultrasound, Lip Videos | English | 7484 samples | Sentence |
AUspeech | 54 | Normal/Dysarthria (Adult) | 920 × 700 | Ultrasound, Audio, Text | Mandarin | 22.31 hours | Vowels, Monosyllable, Sentence, Non-words |
