Tech-
invite
3GPP
space
IETF
space
21
22
23
24
25
26
27
28
29
31
32
33
34
35
36
37
38
4‑5x
Content for
TR 26.943
Word version: 18.0.0
0…
2…
2
References
3
Abbreviations
4
General
5
Recommendation criteria
6
Performance evaluation method
7
Recognition Performance Evaluation Results
A
Key selection phase documents
$
Change history
2
References
p. 5
3
Abbreviations
p. 5
4
General
p. 6
4.1
Project History
p. 6
4.2
Overview of the speech recognition framework for automated voice services work item
p. 8
4.3
Presentation of the following sections
p. 8
5
Recommendation criteria
p. 8
5.1
Overview
p. 8
5.2
Scoring on individual databases
p. 8
5.3
Performance metric over all databases
p. 9
5.4
Comparisons between codecs
p. 9
5.4.1
Low data-rate codec comparison
p. 9
5.4.2
High data-rate codec comparison
p. 9
5.4.2.1
8 kHz sampling rate
p. 9
5.4.2.2
16 kHz sampling rate
p. 9
5.5
Detailed recommendation comparisons
p. 9
6
Performance evaluation method
p. 10
6.1
Introduction
p. 10
6.2
Recognition engines
p. 11
6.2.1
Recognizer for speech codecs based proposals
p. 11
6.2.2
Training and testing
p. 11
6.2.3
Recognizer for DSR
p. 11
6.2.4
Training and testing
p. 11
6.3
Usage of VAD for frame dropping
p. 12
6.4
Codec evaluations
p. 12
6.4.1
Recognition experiments under error-free channel
p. 12
6.5
Recognition experiments under channel errors
p. 14
7
Recognition Performance Evaluation Results
p. 15
A
Key selection phase documents
p. 19
$
Change history
p. 20