Keyphrases
Handset
65%
End-to-end Speech Recognition
63%
Mandarin
62%
Speech Recognition
62%
Taiwan
60%
Taiwanese
59%
Robust Speaker Recognition
44%
Eigen
44%
Prosody Analysis
42%
Mandarin Speech Recognition
41%
Taiwan Mandarin
38%
Formosa
37%
Mismatch Compensation
35%
Error Rate
33%
Speech Corpus
31%
Automatic Speech Recognition
31%
Radio
28%
Slovak
28%
Language Model
28%
Minimum Classification Error
26%
Robust Speaker Verification
26%
Unseen
25%
Prosody
24%
Speech Synthesis
23%
Speech Recognition System
22%
Confidence Measure
22%
Conditional Random Fields
22%
Mandarin Speech
22%
Speaker Identification
21%
Rescoring
20%
Noisy Environment
20%
Deep Learning
19%
Feature Normalization
17%
Dual-microphone
17%
Prosodic Modeling
17%
Robust Speaker Identification
17%
Beamforming
17%
POS Annotation
17%
Taggers
17%
Spectral Features
16%
Speaker Recognition
16%
In the Wild
15%
Maximum Likelihood Linear Regression
15%
Neural Network
15%
Rate Reduction
15%
Utterance
13%
Kaldi
13%
Multi-condition Training
13%
Prosodic Features
13%
Model Weighting
13%
Computer Science
Speech Recognition
100%
Experimental Result
88%
Robust Speech Recognition
63%
Speaker Recognition
57%
Priori Knowledge
39%
Speaker Identification
35%
Deep Learning Method
30%
Classification Error
29%
Confidence Measure
26%
Minimum Classification
26%
Speaker Verification
26%
Large Language Model
26%
maximum-likelihood
25%
Reference Model
21%
Noisy Environment
20%
Speech Synthesis
20%
Language Modeling
20%
Speech Enhancement
20%
Speech Recognition System
19%
Text To Speech
19%
Gaussian Mixture Model
19%
speech corpus
17%
Biometrics
17%
Recurrent Neural Network
17%
Feature Normalization
17%
Transfer Learning
17%
Support Effort
17%
Unknown Speaker
17%
Feature Extraction
16%
Word Error Rate
15%
Histogram Equalization
14%
Systems Performance
14%
Performance Degradation
13%
Telecommunication
13%
Spectral Feature
13%
Recognizer
13%
Recognition Accuracy
13%
Computer Assisted Language Learning
11%
Learning System
11%
User Interface
11%
Neural Network
11%
Document Retrieval
11%
Recognition Rate
11%
Deep Neural Network
11%
Recognition System
11%
Pre-Trained Model
11%
User Experience
10%
Superior Performance
10%
cepstral
9%
Relative Error
9%
