def getAcousticIndices(audiofile):
if (DEBUG_FLAG):
print(
"[WORKING] Attempting to run acoustic indices calculator - acousticIndices.py"
)
# loop through the files in the directory
try:
data, fs = librosa.load(audiofile, sr=None, offset=0, duration=60)
# mono channel
data = AudioProcessing.convert_to_mono(data)
# changing sampling rate
new_fs = 17640
data_chunk = AudioProcessing.resample(data, fs, new_fs)
# extracting indices
acousticIndices = AcousticIndices(data_chunk, new_fs)
acoustic_indices = acousticIndices.get_acoustic_indices()
acoustic_indices = list(map(lambda x: round(x, 4), acoustic_indices))
if (PREDICTION_VERBOSE):
print(acoustic_indices)
acoustic_headers = acousticIndices.get_acoustic_indices_headers()
acoustic_descs = acousticIndices.get_acoustic_indices_descs()
# singleResultArray is used to store the results of one file (List of dictionaries)
singleResultArray = []
# Traverse the acoustic tags
for i in range(len(acoustic_headers)):
# per indices in the length of the acoustic tags,
# append dictionary items.
singleResultArray.append({
"index": acoustic_headers[i],
"value": acoustic_indices[i],
"desc": acoustic_descs[i]
})
# append result dictionary to the final results array
if (DEBUG_FLAG):
print("[WORKING] Calculated " + acoustic_headers[i] +
" - acousticIndices.py")
except Exception as e:
track = traceback.format_exc()
print(track)
singleResultArray = "ERROR_PRESENT"
if (DEBUG_FLAG):
print("[SUCCESS] Calculated acoustic indices - acousticIndices.py")
return singleResultArray
def getAcousticIndices():
# fileDictionary will be used to store the filecount keys with their respective file information
fileDictionary = {}
# Create file counter
fileCount = 0
print(
"[WORKING] Attempting to run acoustic indices calculator - acousticIndices.py"
)
# loop through the files in the directory
for file in os.listdir("instance/upload/"):
# correct the file path with the prefixed upload folder
filePath = "instance/upload/" + file
data, fs = librosa.load(filePath, sr=None, offset=0, duration=60)
# mono channel
data = AudioProcessing.convert_to_mono(data)
# changing sampling rate
new_fs = 17640
data_chunk = AudioProcessing.resample(data, fs, new_fs)
# extracting indices
acousticIndices = AcousticIndices(data_chunk, new_fs)
acoustic_indices = acousticIndices.get_acoustic_indices()
acoustic_headers = acousticIndices.get_acoustic_indices_headers()
# singleResultArray is used to store the results of one file (List of dictionaries)
singleResultArray = []
# Traverse the acoustic tags
for i in range(len(acoustic_headers)):
# per indices in the length of the acoustic tags,
# append dictionary items.
singleResultArray.append({
"index": acoustic_headers[i],
"value": acoustic_indices[i]
})
# append result dictionary to the final results array
print("[WORKING] Calculated " + acoustic_headers[i] +
" - acousticIndices.py")
fileDictionary[fileCount] = singleResultArray
fileCount += 1
print("[SUCCESS] Calculated acoustic indices - acousticIndices.py")
return fileDictionary
return feature_headers
if __name__ == "__main__":
# Audio Filename to be read
filename = "bird.mp3"
# considering one minute of the audio - the indices are taken 1 minute audio segments
data, fs = librosa.load(filename, sr=None, offset=0, duration=60)
# mono channel
data = AudioProcessing.convert_to_mono(data)
# changing sampling rate
new_fs = 17640
data_chunk = AudioProcessing.resample(data, fs, new_fs)
# extracting indices
acousticIndices = AcousticIndices(data_chunk, new_fs)
acoustic_indices = acousticIndices.get_acoustic_indices()
acoustic_headers = acousticIndices.get_acoustic_indices_headers()
acousticIndices = np.column_stack(acousticIndices)
# creating a dataframe
df = pd.DataFrame(acousticIndices)
df.columns = acoustic_headers
df.to_csv("acoustic_indices.csv")