Spoken language identification
Spoken language identification is a subtask in the field of speech processing that focuses on automatically recognizing the language being used from audio input. The task aims to accurately determine the specific language spoken by analyzing the acoustic features of speech signals, thereby providing fundamental support for speech recognition, translation, and interaction in multilingual environments. Its application value lies in enhancing the efficiency and accuracy of cross-language communication, promoting global exchange.
LRE07
VoxForge European
YouTube News dataset (No Noise)
Inception-v3 CRNN
YouTube News dataset (White Noise)
Inception-v3 CRNN
Untranscribed mixed-speech dataset
SVM
VoxForge Commonwealth
IndicTTS
VoxForge
LEAF
VOXLINGUA107
KALAKA-3
YouTube News dataset (Crackling Noise)
Inception-v3 CRNN
YouTube News dataset (Background Music)