def asr_client(
id,
output_file,
input_file,
num_iterations,
simulate_realtime,
jarvis_uri,
max_alternatives,
automatic_punctuation,
word_time_offsets,
):
CHUNK = 1600
channel = grpc.insecure_channel(jarvis_uri)
wf = wave.open(input_file, 'rb')
frames = wf.getnframes()
rate = wf.getframerate()
duration = frames / float(rate)
if id == 0:
print("File duration: %.2fs" % duration)
client = jasr_srv.JarvisASRStub(channel)
config = jasr.RecognitionConfig(
encoding=ja.AudioEncoding.LINEAR_PCM,
sample_rate_hertz=wf.getframerate(),
language_code="en-US",
max_alternatives=max_alternatives,
enable_automatic_punctuation=automatic_punctuation,
enable_word_time_offsets=word_time_offsets,
)
streaming_config = jasr.StreamingRecognitionConfig(
config=config, interim_results=True) # read data
def generator(w, s, num_iterations, output_file):
try:
for i in range(num_iterations):
w = wave.open(input_file, 'rb')
start_time = time.time()
yield jasr.StreamingRecognizeRequest(streaming_config=s)
num_requests = 0
while 1:
d = w.readframes(CHUNK)
if len(d) <= 0:
break
num_requests += 1
if simulate_realtime:
time_to_sleep = max(
0.0, CHUNK / rate * num_requests -
(time.time() - start_time))
time.sleep(time_to_sleep)
yield jasr.StreamingRecognizeRequest(audio_content=d)
w.close()
except Exception as e:
print(e)
responses = client.StreamingRecognize(
generator(wf, streaming_config, num_iterations, output_file))
print_to_file(responses, output_file, max_alternatives, word_time_offsets)
def main():
args = get_args()
if args.list_devices:
p = pyaudio.PyAudio()
for i in range(p.get_device_count()):
info = p.get_device_info_by_index(i)
if info['maxInputChannels'] < 1:
continue
print(f"{info['index']}: {info['name']}")
sys.exit(0)
channel = grpc.insecure_channel(args.server)
client = jasr_srv.JarvisASRStub(channel)
config = jasr.RecognitionConfig(
encoding=ja.AudioEncoding.LINEAR_PCM,
sample_rate_hertz=RATE,
language_code="en-US",
max_alternatives=1,
enable_automatic_punctuation=True,
)
streaming_config = jasr.StreamingRecognitionConfig(config=config,
interim_results=True)
with MicrophoneStream(RATE, CHUNK, device=args.input_device) as stream:
audio_generator = stream.generator()
requests = (jasr.StreamingRecognizeRequest(audio_content=content)
for content in audio_generator)
def build_generator(cfg, gen):
yield jasr.StreamingRecognizeRequest(streaming_config=cfg)
for x in gen:
yield x
responses = client.StreamingRecognize(
build_generator(streaming_config, requests))
listen_print_loop(responses)
import jarvis_api.jarvis_nlp_pb2_grpc as jnlp_srv
# ASR proto
import jarvis_api.jarvis_asr_pb2 as jasr
import jarvis_api.jarvis_asr_pb2_grpc as jasr_srv
# TTS proto
import jarvis_api.jarvis_tts_pb2 as jtts
import jarvis_api.jarvis_tts_pb2_grpc as jtts_srv
import jarvis_api.audio_pb2 as ja
import json
app = Flask(__name__)
channel = grpc.insecure_channel('localhost:50051')
jarvis_asr = jasr_srv.JarvisASRStub(channel)
jarvis_nlp = jnlp_srv.JarvisNLPStub(channel)
jarvis_tts = jtts_srv.JarvisTTSStub(channel)
def TTS(text):
req = jtts.SynthesizeSpeechRequest()
req.text = text
req.language_code = "en-US" # currently required to be "en-US"
req.encoding = ja.AudioEncoding.LINEAR_PCM # Supports LINEAR_PCM, FLAC, MULAW and ALAW audio encodings
req.sample_rate_hz = 22050 # ignored, audio returned will be 22.05KHz
req.voice_name = "ljspeech" # ignored
resp = jarvis_tts.Synthesize(req)
audio_samples = np.frombuffer(resp.audio, dtype=np.float32)
return {'content': audio_samples.tolist(), 'sr': 22050}
default="localhost:50051",
type=str,
help="URI to GRPC server endpoint")
parser.add_argument("--audio-file",
required=True,
help="path to local file to stream")
return parser.parse_args()
args = get_args()
wf = wave.open(args.audio_file, 'rb')
with open(args.audio_file, 'rb') as fh:
data = fh.read()
channel = grpc.insecure_channel(args.server)
client = jasr_srv.JarvisASRStub(channel)
config = jasr.RecognitionConfig(
encoding=ja.AudioEncoding.LINEAR_PCM,
sample_rate_hertz=wf.getframerate(),
language_code="en-US",
max_alternatives=1,
enable_automatic_punctuation=False,
audio_channel_count=1,
)
request = jasr.RecognizeRequest(config=config, audio=data)
response = client.Recognize(request)
print(response)