HMM-TDNN(LFMMI) + LSTMLM + NN-GEV

Reported on 1 benchmark across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Audio1 result

Speech RecognitiononCHiME-4 real 6ch
Word Error Rate (WER)· 2018-03-27
2.74
best: 1.99 (Complex Spectral Mapping + WRBN + Utterance-Wise Dropout + Iterative Speaker Adaptation)
SOTA
Building state-of-the-art distant speech recognition using the CHiME-4 challenge with a setup of speech enhancement baseline arXiv:1803.10109