Addition of Noise Suppression and Gain Control to Profiles (spokestack#37)

will-rice · web-flow · commit fc0797b1a644 · 2021-01-25T08:38:14.000-05:00
diff --git a/spokestack/profile/vad_trigger_asr.py b/spokestack/profile/vad_trigger_asr.py
@@ -4,8 +4,10 @@
 from typing import Any
 
 from spokestack.activation_timeout import ActivationTimeout
+from spokestack.agc.webrtc import AutomaticGainControl
 from spokestack.asr.spokestack.speech_recognizer import CloudSpeechRecognizer
 from spokestack.io.pyaudio import PyAudioInput
+from spokestack.nsx.webrtc import AutomaticNoiseSuppression
 from spokestack.pipeline import SpeechPipeline
 from spokestack.vad.webrtc import VoiceActivityDetector, VoiceActivityTrigger
 
@@ -38,6 +40,8 @@ def create(
                 sample_rate=sample_rate, frame_width=frame_width, **kwargs
             ),
             stages=[
+                AutomaticGainControl(sample_rate=sample_rate, frame_width=frame_width),
+                AutomaticNoiseSuppression(sample_rate=sample_rate),
                 VoiceActivityDetector(
                     sample_rate=sample_rate, frame_width=frame_width, **kwargs
                 ),
diff --git a/spokestack/profile/wakeword_asr.py b/spokestack/profile/wakeword_asr.py
@@ -4,8 +4,10 @@
 from typing import Any
 
 from spokestack.activation_timeout import ActivationTimeout
+from spokestack.agc.webrtc import AutomaticGainControl
 from spokestack.asr.spokestack.speech_recognizer import CloudSpeechRecognizer
 from spokestack.io.pyaudio import PyAudioInput
+from spokestack.nsx.webrtc import AutomaticNoiseSuppression
 from spokestack.pipeline import SpeechPipeline
 from spokestack.vad.webrtc import VoiceActivityDetector
 from spokestack.wakeword.tflite import WakewordTrigger
@@ -40,6 +42,8 @@ def create(
                 frame_width=frame_width, sample_rate=sample_rate, **kwargs
             ),
             stages=[
+                AutomaticGainControl(sample_rate=sample_rate, frame_width=frame_width),
+                AutomaticNoiseSuppression(sample_rate=sample_rate),
                 VoiceActivityDetector(
                     frame_width=frame_width,
                     sample_rate=sample_rate,
