def __init__(self, args):
super().__init__(args)
self.listener = PocketsphinxListener(args.key_phrase, args.dict_file,
args.hmm_folder, args.threshold)
self.outputs = []
self.targets = []
self.filenames = []
def main():
args = TrainData.parse_args(create_parser(usage))
data = TrainData.from_both(args.tags_file, args.tags_folder, args.folder)
data_files = data.train_files if args.use_train else data.test_files
listener = PocketsphinxListener(args.key_phrase, args.dict_file,
args.hmm_folder, args.threshold)
print('Data:', data)
stats = test_pocketsphinx(listener, data_files)
show_stats(stats, not args.no_filenames)
def main():
parser = create_parser(usage)
parser.add_argument(
'models',
nargs='*',
help='List of model filenames in format: wake-word.yy-mm-dd.net')
args = TrainData.parse_args(parser)
if not (bool(args.pocketsphinx_dict) == bool(args.pocketsphinx_folder) ==
bool(args.pocketsphinx_wake_word)):
parser.error('Must pass all or no Pocketsphinx arguments')
data = TrainData.from_both(args.tags_file, args.tags_folder, args.folder)
data_files = data.train_files if args.use_train else data.test_files
print('Data:', data)
metrics = {}
if args.pocketsphinx_dict and args.pocketsphinx_folder and args.pocketsphinx_wake_word:
if not isfile(args.pocketsphinx_dict):
parser.error('No such file: ' + args.pocketsphinx_dict)
if not isdir(args.pocketsphinx_folder):
parser.error('No such folder: ' + args.pocketsphinx_folder)
listener = PocketsphinxListener(args.pocketsphinx_wake_word,
args.pocketsphinx_dict,
args.pocketsphinx_folder,
args.pocketsphinx_threshold)
stats = test_pocketsphinx(listener, data_files)
metrics[args.pocketsphinx_dict] = stats_to_dict(stats)
for model_name in args.models:
print('Calculating', model_name + '...')
inject_params(model_name)
train, test = data.load(args.use_train, not args.use_train)
inputs, targets = train if args.use_train else test
predictions = Listener.find_runner(model_name)(model_name).predict(
inputs)
stats = Stats(predictions, targets, sum(data_files, []))
print('----', model_name, '----')
print(stats.counts_str())
print()
print(stats.summary_str())
print()
metrics[model_name] = stats.to_dict(args.threshold)
print('Writing to:', args.output)
with open(args.output, 'w') as f:
json.dump(metrics, f)
def main():
args = create_parser(usage).parse_args()
def on_activation():
Popen(['aplay', '-q', 'data/activate.wav'])
def on_prediction(conf):
print('!' if conf > 0.5 else '.', end='', flush=True)
runner = PreciseRunner(ListenerEngine(
PocketsphinxListener(args.key_phrase, args.dict_file, args.hmm_folder,
args.threshold, args.chunk_size)),
3,
on_activation=on_activation,
on_prediction=on_prediction)
runner.start()
Event().wait() # Wait forever
def run(self):
def on_activation():
activate_notify()
def on_prediction(conf):
print('!' if conf > 0.5 else '.', end='', flush=True)
args = self.args
runner = PreciseRunner(ListenerEngine(
PocketsphinxListener(args.key_phrase, args.dict_file,
args.hmm_folder, args.threshold,
args.chunk_size)),
3,
on_activation=on_activation,
on_prediction=on_prediction)
runner.start()
Event().wait() # Wait forever
class PocketsphinxTestScript(BaseScript):
usage = Usage('''
Test a dataset using Pocketsphinx
:key_phrase str
Key phrase composed of words from dictionary
:dict_file str
Filename of dictionary with word pronunciations
:hmm_folder str
Folder containing hidden markov model
:-th --threshold str 1e-90
Threshold for activations
:-t --use-train
Evaluate training data instead of test data
:-nf --no-filenames
Don't show the names of files that failed
...
''') | TrainData.usage
def __init__(self, args):
super().__init__(args)
self.listener = PocketsphinxListener(args.key_phrase, args.dict_file,
args.hmm_folder, args.threshold)
self.outputs = []
self.targets = []
self.filenames = []
def get_stats(self):
return Stats(self.outputs, self.targets, self.filenames)
def run(self):
args = self.args
data = TrainData.from_both(args.tags_file, args.tags_folder,
args.folder)
print('Data:', data)
ww_files, nww_files = data.train_files if args.use_train else data.test_files
self.run_test(ww_files, 'Wake Word', 1.0)
self.run_test(nww_files, 'Not Wake Word', 0.0)
stats = self.get_stats()
if not self.args.no_filenames:
fp_files = stats.calc_filenames(False, True, 0.5)
fn_files = stats.calc_filenames(False, False, 0.5)
print('=== False Positives ===')
print('\n'.join(fp_files))
print()
print('=== False Negatives ===')
print('\n'.join(fn_files))
print()
print(stats.counts_str(0.5))
print()
print(stats.summary_str(0.5))
def eval_file(self, filename) -> float:
transcription = check_output([
'pocketsphinx_continuous', '-kws_threshold', '1e-20', '-keyphrase',
'hey my craft', '-infile', filename
],
stderr=PIPE)
return float(bool(transcription) and not transcription.isspace())
def run_test(self, test_files, label_name, label):
print()
print('===', label_name, '===')
for test_file in test_files:
try:
with wave.open(test_file) as wf:
frames = wf.readframes(wf.getnframes())
except (OSError, EOFError):
print('?', end='', flush=True)
continue
out = int(self.listener.found_wake_word(frames))
self.outputs.append(out)
self.targets.append(label)
self.filenames.append(test_file)
print('!' if out else '.', end='', flush=True)
print()