birbnetes/cnn-classification-service
birbnetes/cnn-classification-service
4
0
Fork 0
Code Pull Requests Releases Wiki Activity

tracing fixes
All checks were successful
continuous-integration/drone/push Build is passing
Details

Browse Source
This commit is contained in:
Pünkösd Marcell 2021-08-04 15:20:29 +02:00
parent 977fc98ed2
commit ee9dea06fb
3 changed files with 34 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
cnn_classification_service/cnn_classifier.py
View File

@ -29,28 +29,32 @@ class Classifier(object):
@staticmethod
def create_spectrogram(wav_filename: str) -> Tuple[str, str]:
matplotlib.pyplot.interactive(False)
clip, sample_rate = librosa.load(wav_filename, sr=None)
fig = matplotlib.pyplot.figure(figsize=[0.72, 0.72])
ax = fig.add_subplot(111)
ax.axes.get_xaxis().set_visible(False)
ax.axes.get_yaxis().set_visible(False)
ax.set_frame_on(False)
spectogram = librosa.feature.melspectrogram(y=clip, sr=sample_rate)
librosa.display.specshow(librosa.power_to_db(spectogram, ref=numpy.max))
target_dir = tempfile.mkdtemp(dir="/dev/shm")
wav_basename = os.path.basename(wav_filename)
with opentracing.tracer.start_active_span('classifier.librosa.load'):
clip, sample_rate = librosa.load(wav_filename, sr=None)
# Change extension to jpg... mert 110% biztos vagyok benne hogy a keras nem bírná beolvasni máshogy
file_name = os.path.join(target_dir, "unknown", f"{wav_basename[:-4]}.jpg")
os.mkdir(os.path.join(target_dir, "unknown"))
with opentracing.tracer.start_active_span('classifier.Plot'):
matplotlib.pyplot.interactive(False)
fig = matplotlib.pyplot.figure(figsize=[0.72, 0.72])
ax = fig.add_subplot(111)
ax.axes.get_xaxis().set_visible(False)
ax.axes.get_yaxis().set_visible(False)
ax.set_frame_on(False)
spectogram = librosa.feature.melspectrogram(y=clip, sr=sample_rate)
librosa.display.specshow(librosa.power_to_db(spectogram, ref=numpy.max))
matplotlib.pyplot.savefig(file_name, dpi=400, bbox_inches='tight', pad_inches=0)
matplotlib.pyplot.close()
fig.clf()
matplotlib.pyplot.close(fig)
matplotlib.pyplot.close('all')
target_dir = tempfile.mkdtemp(dir="/dev/shm")
wav_basename = os.path.basename(wav_filename)
# Change extension to jpg... mert 110% biztos vagyok benne hogy a keras nem bírná beolvasni máshogy
file_name = os.path.join(target_dir, "unknown", f"{wav_basename[:-4]}.jpg")
os.mkdir(os.path.join(target_dir, "unknown"))
matplotlib.pyplot.savefig(file_name, dpi=400, bbox_inches='tight', pad_inches=0)
matplotlib.pyplot.close()
fig.clf()
matplotlib.pyplot.close(fig)
matplotlib.pyplot.close('all')
return target_dir, file_name # Az unknown nélkülivel kell visszatérni
@ -85,12 +89,10 @@ class Classifier(object):
return predicted_class_name, labeled_predictions
def predict(self, wav_filename: str) -> Tuple[str, dict]:
span = opentracing.tracer.scope_manager.active().span
with opentracing.tracer.start_active_span('createSpectrogram', child_of=span):
with opentracing.tracer.start_active_span('classifier.createSpectrogram'):
directory, _ = self.create_spectrogram(wav_filename)
with opentracing.tracer.start_active_span('runPredictor', child_of=span):
with opentracing.tracer.start_active_span('classifier.runPredictor'):
result = self._run_predictor(directory)
shutil.rmtree(directory) # The image is no longer needed

10
cnn_classification_service/magic_doer.py
View File

@ -19,15 +19,16 @@ class MagicDoer:
@classmethod
def run_everything(cls, parameters: dict) -> dict:
span = opentracing.tracer.scope_manager.active().span
tag = parameters['tag']
sample_file_handle, sample_file_path = tempfile.mkstemp(prefix=f"{tag}_", suffix=".wav", dir="/dev/shm")
span = opentracing.tracer.scope_manager.active().span
span.log_kv({'event': 'sampleFileOpened', 'sampleTag': tag})
response = None
try:
with opentracing.tracer.start_active_span('downloadSample', child_of=span):
with opentracing.tracer.start_active_span('magicDoer.downloadSample'):
# Download Sample
object_path = urljoin(Config.STORAGE_SERVICE_URL, f"object/{tag}")
@ -38,11 +39,11 @@ class MagicDoer:
logging.debug(f"Downloaded sample to {sample_file_path}")
with opentracing.tracer.start_active_span('loadClassifier', child_of=span):
with opentracing.tracer.start_active_span('magicDoer.getClassifier'):
# Get a classifier that uses the default model
model_details, classifier = cls.classifier_cache.get_default_classifier()
with opentracing.tracer.start_active_span('runClassifier', child_of=span) as child_span:
with opentracing.tracer.start_active_span('magicDoer.runClassifier'):
# do the majic
classification_start_time = time.time()
predicted_class_name, labeled_predictions = classifier.predict(sample_file_path)
@ -63,6 +64,7 @@ class MagicDoer:
except FileNotFoundError:
pass
span = opentracing.tracer.scope_manager.active().span
span.log_kv({'event': 'sampleFileDeleted', 'sampleTag': tag})
if not response:

6
cnn_classification_service/main.py
View File

@ -19,16 +19,16 @@ from magic_doer import MagicDoer
def message_callback(channel, method, properties, body):
with opentracing.tracer.start_active_span('messageHandling', finish_on_close=True) as span:
with opentracing.tracer.start_active_span('main.handleMessage', finish_on_close=True) as scope:
try:
msg = json.loads(body.decode('utf-8'))
except (UnicodeDecodeError, json.JSONDecodeError) as e:
logging.warning(f"Invalid message recieved: {e}")
return
span.log_kv({'event': 'messageParsed', 'sampleTag': msg['tag']})
scope.span.log_kv({'event': 'messageParsed', 'sampleTag': msg['tag']})
with opentracing.tracer.start_active_span('runAlgorithm', child_of=span) as child_span:
with opentracing.tracer.start_active_span('magicDoer.run_everything'):
results = MagicDoer.run_everything(msg) # <- This is where the magic happens
if results: