def predict_url(urls, merge=True):
"""
Function to predict an url
"""
catch_url_error(urls)
if not loaded:
load_inference_model()
urllib.request.urlretrieve(urls['urls'][0], '/tmp/file.wav')
pred_lab, pred_prob =label_wav.predict('/tmp/file.wav', LABELS_FILE, MODEL_NAME, "wav_data:0","labels_softmax:0", 3)
return format_prediction(pred_lab, pred_prob)
def predict_data(audios, merge=True):
"""
Function to predict an audio file
"""
if not loaded:
load_inference_model()
if not isinstance(audios, list):
audios = [audios]
filenames = []
for audio in audios:
thename=audio['files'].filename
thefile="/tmp/"+thename
audio['files'].save(thefile)
pred_lab, pred_prob =label_wav.predict(thefile, LABELS_FILE, MODEL_NAME, "wav_data:0","labels_softmax:0", 3)
return format_prediction(pred_lab, pred_prob)
def predict_data(args):
"""
Function to predict an audio file
"""
# # Check user configuration
# update_with_query_conf(args)
# conf = config.conf_dict
if not loaded:
load_inference_model()
# Create a list with the path to the audios
filenames = [f.filename for f in args['files']]
pred_lab, pred_prob = label_wav.predict(filenames[0], LABELS_FILE,
MODEL_NAME, "wav_data:0",
"labels_softmax:0", 3)
return format_prediction(pred_lab, pred_prob)
def predict_url(args):
"""
Function to predict an url
"""
# # Check user configuration
# update_with_query_conf(args)
# conf = config.conf_dict
catch_url_error(args['urls'])
# Load model if needed
if not loaded:
load_inference_model()
# Download the url
urllib.request.urlretrieve(args['urls'][0], '/tmp/file.wav')
pred_lab, pred_prob = label_wav.predict('/tmp/file.wav', LABELS_FILE,
MODEL_NAME, "wav_data:0",
"labels_softmax:0", 3)
return format_prediction(pred_lab, pred_prob)