Speech Recognition on CHiME real

Metric: Percentage error (lower is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Hide extra data

Sort:

#	Model↕	Percentage error▲	Extra Data	Paper	Date↕	Code
1	Distortion-Independent + WRBN + Utterance-Wise Recurrent Dropout + Magnitude Features	7.8	Yes	-	-	-
2	HMM-TDNN(LFMMI) + LSTMLM	11.4	No	Building state-of-the-art distant speech recogni...	2018-03-27	-
3	Li-GRU	14.6	No	The PyTorch-Kaldi Speech Recognition Toolkit	2018-11-19	Code
4	Deep Speech 2	21.79	No	Deep Speech 2: End-to-End Speech Recognition in ...	2015-12-08	Code
5	CNN + Bi-RNN + CTC (speech to letters)	67.94	No	Deep Speech: Scaling up end-to-end speech recogn...	2014-12-17	Code

#1Distortion-Independent + WRBN + Utterance-Wise Recurrent Dropout + Magnitude Features
7.8
Percentage error· Extra Data
No paper
#2HMM-TDNN(LFMMI) + LSTMLM SOTA
11.4
Percentage error· 2018-03-27
Building state-of-the-art distant speech recognition using the CHiME-4 challenge with a setup of speech enhancement baseline
#3Li-GRU
14.6
Percentage error· 2018-11-19
The PyTorch-Kaldi Speech Recognition Toolkit Code
#4Deep Speech 2SOTA
21.79
Percentage error· 2015-12-08
Deep Speech 2: End-to-End Speech Recognition in English and Mandarin Code
#5CNN + Bi-RNN + CTC (speech to letters)SOTA
67.94
Percentage error· 2014-12-17
Deep Speech: Scaling up end-to-end speech recognition Code