-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathtask3_sound_event_detection_in_real_life_audio.yaml
80 lines (67 loc) · 2.3 KB
/
task3_sound_event_detection_in_real_life_audio.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
# ==========================================================
# Flow
# ==========================================================
flow:
initialize: true
extract_features: true
feature_normalizer: true
train_system: true
test_system: true
evaluate_system: true
# ==========================================================
# General
# ==========================================================
general:
development_dataset: TUTSoundEvents_2016_DevelopmentSet
challenge_dataset: TUTSoundEvents_2016_EvaluationSet
overwrite: false
# ==========================================================
# Paths
# ==========================================================
path:
data: data/
base: system/baseline_dcase2016_task3/
features: features/
feature_normalizers: feature_normalizers/
models: acoustic_models/
results: evaluation_results/
challenge_results: challenge_submission/task_3_sound_event_detection_in_real_life_audio/
# ==========================================================
# Feature extraction
# ==========================================================
features:
fs: 44100
win_length_seconds: 0.04
hop_length_seconds: 0.02
include_mfcc0: false
include_delta: true
include_acceleration: true
mfcc:
n_mfcc: 20 # Number of MFCC coefficients
n_mels: 40 # Number of MEL bands used
n_fft: 2048 # FFT length, window length in samples
fmin: 0 # Minimum frequency when constructing MEL bands
fmax: 22050 # Maximum frequency when constructing MEL band
mfcc_delta:
width: 9
mfcc_acceleration:
width: 9
# ==========================================================
# Classifier
# ==========================================================
classifier:
method: gmm # Currently system supports only gmm
parameters: !!null # Parameters are copied from classifier.parameters based on defined method
classifier_parameters:
gmm:
n_components: 16
min_covar: 0.001
n_iter: 40
# ==========================================================
# Detector
# ==========================================================
detector:
decision_threshold: 120.0
smoothing_window_length: 1.0
minimum_event_length: 0.1
minimum_event_gap: 0.1