def waveform():
filename = request.args.get('filename')
name = extractFilename(filename)
outputpath = "output/" + str(name) + "/"
mkDIR(outputpath)
status = wave.audio_waveplot(INPUTPATH=filename, OUTPATH=outputpath)
return status
def analyze():
filename = request.args.get('filename')
name = extractFilename(filename)
outputpath = "output/" + str(name) + "/"
mkDIR(outputpath)
status = zcr_sigenergy(INPUTPATH=filename, OUTPATH=outputpath)
return status
def transcribe_data():
filename = request.args.get('filename')
name = extractFilename(filename)
data = {}
path = "output/" + str(name) + "/"
with open(path + 'transcribe.json', 'r') as transcribe:
data = transcribe.read()
return data
def fe():
filename = request.args.get('filename')
name = extractFilename(filename)
outputpath = "output/" + str(name) + "/"
mkDIR(outputpath)
status = feature_extraction(INPUTPATH=filename, OUTPATH=outputpath)
return status
def transcribe():
AUDIO_FILE = request.args.get('filename')
with open(AUDIO_FILE, 'rb') as f:
resp = client.speech(f, None, {'Content-Type': 'audio/wav'})
text = resp["_text"]
word_frequency = {}
for word in text.split():
if word in word_frequency.keys():
word_frequency[word] = word_frequency[word] + 1
else:
word_frequency[word] = 1
resp.update({'word_frequency': word_frequency})
name = extractFilename(AUDIO_FILE)
outputpath = "output/" + str(name) + "/"
with open(outputpath + 'transcribe.json', 'w') as transcribe:
transcribe.write(json.dumps(resp))
return (json.jsonify(resp))
def mysptotal(sound, sourcerun):
path, filename = extractFilename(sound, withextension=True)
path = path + "/"
try:
objects = parselmouth.praat.run_file(sourcerun,
-20,
2,
0.3,
"yes",
sound,
path,
80,
400,
0.01,
capture_output=True)
#print (objects[0]) # This will print the info from the sound object, and objects[0] is a parselmouth.Sound object
z1 = str(
objects[1]
) # This will print the info from the textgrid object, and objects[1] is a parselmouth.Data object with a TextGrid inside
z2 = z1.strip().split()
z3 = np.array(z2)
z4 = np.array(z3)[np.newaxis]
z5 = z4.T
db = binom.rvs(n=10, p=float(z2[14]), size=10000)
a = np.array(db)
b = np.mean(a) * 100 / 10
dataset = pd.DataFrame({
"number_of_syllables":
z5[0, :],
"number_of_pauses":
z5[1, :],
"rate_of_speech":
z5[2, :],
"articulation_rate":
z5[3, :],
"speaking_duration":
z5[4, :],
"original_duration":
z5[5, :],
"pronunciation_posteriori_probability_score_percentage":
b,
"balance":
z5[6, :],
"f0_mean":
z5[7, :],
"f0_std":
z5[8, :],
"f0_median":
z5[9, :],
"f0_min":
z5[10, :],
"f0_max":
z5[11, :],
"f0_quantile25":
z5[12, :],
"f0_quan75":
z5[13, :]
})
return dataset.to_json(orient='index')
except Exception as e:
return {"Try again the sound of the audio was not clear" + str(e)}
def wave_image():
filename = request.args.get('filename')
name = extractFilename(filename)
return send_file('output/' + name + '/waveplot.png',
attachment_filename='waveplot.png',
mimetype='image/png')
def analyze_image():
filename = request.args.get('filename')
name = extractFilename(filename)
return send_file('output/' + name + '/zcr_energy.png',
attachment_filename='zcr_energy.png',
mimetype='image/png')