def _append_trivial_tree(self, text_file, sync_root):
"""
Append trivial tree, made by one HEAD,
one sync map fragment for each element of ``text_file``,
and one TAIL.
This function is called if either ``text_file`` has only one element,
or if ``sync_root.value`` is an interval with zero length
(i.e., ``sync_root.value.begin == sync_root.value.end``).
"""
interval = sync_root.value
#
# NOTE the following is correct, but it is a bit obscure
# time_values = [interval.begin] * (1 + len(text_file)) + [interval.end] * 2
#
if len(text_file) == 1:
time_values = [interval.begin, interval.begin, interval.end, interval.end]
else:
# interval.begin == interval.end
time_values = [interval.begin] * (3 + len(text_file))
aba = AdjustBoundaryAlgorithm(rconf=self.rconf, logger=self.logger)
aba.intervals_to_fragment_list(
text_file=text_file,
time_values=time_values
)
aba.append_fragment_list_to_sync_root(sync_root=sync_root)
def _adjust_boundaries(self, boundary_indices, text_file, real_wave_mfcc, sync_root, force_aba_auto=False, leaf_level=False):
"""
Adjust boundaries as requested by the user.
Return the computed time map, that is,
a list of pairs ``[start_time, end_time]``,
of length equal to number of fragments + 2,
where the two extra elements are for
the HEAD (first) and TAIL (last).
"""
# boundary_indices contains the boundary indices in the all_mfcc of real_wave_mfcc
# starting with the (head-1st fragment) and ending with (-1th fragment-tail)
aba_parameters = self.task.configuration.aba_parameters()
if force_aba_auto:
self.log(u"Forced running algorithm: 'auto'")
aba_parameters["algorithm"] = (AdjustBoundaryAlgorithm.AUTO, [])
# note that the other aba settings (nonspeech and nozero)
# remain as specified by the user
self.log([u"ABA parameters: %s", aba_parameters])
aba = AdjustBoundaryAlgorithm(rconf=self.rconf, logger=self.logger)
aba.adjust(
aba_parameters=aba_parameters,
real_wave_mfcc=real_wave_mfcc,
boundary_indices=boundary_indices,
text_file=text_file,
allow_arbitrary_shift=leaf_level
)
aba.append_fragment_list_to_sync_root(sync_root=sync_root)
def _adjust_boundaries(self, text_map, real_wave_full_mfcc,
real_wave_length):
"""
Adjust the boundaries between consecutive fragments.
Return a pair:
1. a success bool flag
2. the computed interval map, that is,
a list of triples ``[start_time, end_time, fragment_id]``
"""
self._log("Adjusting boundaries")
algo = self.task.configuration.adjust_boundary_algorithm
value = None
if algo is None:
self._log("No adjust boundary algorithm specified: returning")
return (True, text_map)
elif algo == AdjustBoundaryAlgorithm.AUTO:
self._log("Requested adjust boundary algorithm AUTO: returning")
return (True, text_map)
elif algo == AdjustBoundaryAlgorithm.AFTERCURRENT:
value = self.task.configuration.adjust_boundary_aftercurrent_value
elif algo == AdjustBoundaryAlgorithm.BEFORENEXT:
value = self.task.configuration.adjust_boundary_beforenext_value
elif algo == AdjustBoundaryAlgorithm.OFFSET:
value = self.task.configuration.adjust_boundary_offset_value
elif algo == AdjustBoundaryAlgorithm.PERCENT:
value = self.task.configuration.adjust_boundary_percent_value
elif algo == AdjustBoundaryAlgorithm.RATE:
value = self.task.configuration.adjust_boundary_rate_value
elif algo == AdjustBoundaryAlgorithm.RATEAGGRESSIVE:
value = self.task.configuration.adjust_boundary_rate_value
self._log(["Requested algo %s and value %s", algo, value])
try:
self._log("Running VAD...")
vad = VAD(logger=self.logger)
vad.wave_mfcc = real_wave_full_mfcc
vad.wave_len = real_wave_length
vad.compute_vad()
self._log("Running VAD... done")
except Exception as e:
self._log("Adjusting boundaries: failed")
self._log(["Message: %s", str(e)])
return (False, None)
self._log("Creating AdjustBoundaryAlgorithm object")
adjust_boundary = AdjustBoundaryAlgorithm(algorithm=algo,
text_map=text_map,
speech=vad.speech,
nonspeech=vad.nonspeech,
value=value,
logger=self.logger)
self._log("Adjusting boundaries...")
adjusted_map = adjust_boundary.adjust()
self._log("Adjusting boundaries... done")
self._log("Adjusting boundaries: succeeded")
return (True, adjusted_map)
def run_aba(self, algorithm, value, expected):
aba = AdjustBoundaryAlgorithm(
algorithm=algorithm,
text_map=self.TEXT_MAP,
speech=self.SPEECH,
nonspeech=self.NONSPEECH,
value=value
)
adjusted_map = aba.adjust()
self.assertEqual(self.maps_are_equal(adjusted_map, self.TEXT_MAP), expected)
def _adjust_boundaries(self,
real_wave_mfcc,
text_file,
boundary_indices,
adjust_boundaries=True):
"""
Adjust boundaries as requested by the user.
Return the computed time map, that is,
a list of pairs ``[start_time, end_time]``,
of length equal to number of fragments + 2,
where the two extra elements are for
the HEAD (first) and TAIL (last).
"""
# boundary_indices contains the boundary indices in the all_mfcc of real_wave_mfcc
# starting with the (head-1st fragment) and ending with (-1th fragment-tail)
if adjust_boundaries:
aba_algorithm, aba_parameters = self.task.configuration.aba_parameters(
)
self.log([u"Running algorithm: '%s'", aba_algorithm])
else:
self.log(u"Forced running algorithm: 'auto'")
aba_algorithm = AdjustBoundaryAlgorithm.AUTO
aba_parameters = None
return AdjustBoundaryAlgorithm(algorithm=aba_algorithm,
parameters=aba_parameters,
real_wave_mfcc=real_wave_mfcc,
boundary_indices=boundary_indices,
text_file=text_file,
rconf=self.rconf,
logger=self.logger).to_time_map()
def _append_trivial_tree(self, text_file, sync_root):
"""
Append trivial tree, made by one HEAD,
one sync map fragment for each element of ``text_file``,
and one TAIL.
This function is called if either ``text_file`` has only one element,
or if ``sync_root.value`` is an interval with zero length
(i.e., ``sync_root.value.begin == sync_root.value.end``).
"""
interval = sync_root.value
#
# NOTE the following is correct, but it is a bit obscure
# time_values = [interval.begin] * (1 + len(text_file)) + [interval.end] * 2
#
if len(text_file) == 1:
time_values = [
interval.begin, interval.begin, interval.end, interval.end
]
else:
# interval.begin == interval.end
time_values = [interval.begin] * (3 + len(text_file))
aba = AdjustBoundaryAlgorithm(rconf=self.rconf, logger=self.logger)
aba.intervals_to_fragment_list(text_file=text_file,
time_values=time_values)
aba.append_fragment_list_to_sync_root(sync_root=sync_root)
def _adjust_boundaries(self,
boundary_indices,
text_file,
real_wave_mfcc,
sync_root,
force_aba_auto=False,
leaf_level=False):
"""
Adjust boundaries as requested by the user.
Return the computed time map, that is,
a list of pairs ``[start_time, end_time]``,
of length equal to number of fragments + 2,
where the two extra elements are for
the HEAD (first) and TAIL (last).
"""
# boundary_indices contains the boundary indices in the all_mfcc of real_wave_mfcc
# starting with the (head-1st fragment) and ending with (-1th fragment-tail)
aba_parameters = self.task.configuration.aba_parameters()
if force_aba_auto:
self.log(u"Forced running algorithm: 'auto'")
aba_parameters["algorithm"] = (AdjustBoundaryAlgorithm.AUTO, [])
# note that the other aba settings (nonspeech and nozero)
# remain as specified by the user
self.log([u"ABA parameters: %s", aba_parameters])
aba = AdjustBoundaryAlgorithm(rconf=self.rconf, logger=self.logger)
aba.adjust(aba_parameters=aba_parameters,
real_wave_mfcc=real_wave_mfcc,
boundary_indices=boundary_indices,
text_file=text_file,
allow_arbitrary_shift=leaf_level)
aba.append_fragment_list_to_sync_root(sync_root=sync_root)
def _adjust_boundaries(
self,
text_map,
real_wave_full_mfcc,
real_wave_length
):
"""
Adjust the boundaries between consecutive fragments.
Return a pair:
1. a success bool flag
2. the computed interval map, that is,
a list of triples ``[start_time, end_time, fragment_id]``
"""
self._log("Adjusting boundaries")
algo = self.task.configuration.adjust_boundary_algorithm
value = None
if algo is None:
self._log("No adjust boundary algorithm specified: returning")
return (True, text_map)
elif algo == AdjustBoundaryAlgorithm.AUTO:
self._log("Requested adjust boundary algorithm AUTO: returning")
return (True, text_map)
elif algo == AdjustBoundaryAlgorithm.AFTERCURRENT:
value = self.task.configuration.adjust_boundary_aftercurrent_value
elif algo == AdjustBoundaryAlgorithm.BEFORENEXT:
value = self.task.configuration.adjust_boundary_beforenext_value
elif algo == AdjustBoundaryAlgorithm.OFFSET:
value = self.task.configuration.adjust_boundary_offset_value
elif algo == AdjustBoundaryAlgorithm.PERCENT:
value = self.task.configuration.adjust_boundary_percent_value
elif algo == AdjustBoundaryAlgorithm.RATE:
value = self.task.configuration.adjust_boundary_rate_value
elif algo == AdjustBoundaryAlgorithm.RATEAGGRESSIVE:
value = self.task.configuration.adjust_boundary_rate_value
self._log(["Requested algo %s and value %s", algo, value])
try:
self._log("Running VAD...")
vad = VAD(logger=self.logger)
vad.wave_mfcc = real_wave_full_mfcc
vad.wave_len = real_wave_length
vad.compute_vad()
self._log("Running VAD... done")
except Exception as e:
self._log("Adjusting boundaries: failed")
self._log(["Message: %s", str(e)])
return (False, None)
self._log("Creating AdjustBoundaryAlgorithm object")
adjust_boundary = AdjustBoundaryAlgorithm(
algorithm=algo,
text_map=text_map,
speech=vad.speech,
nonspeech=vad.nonspeech,
value=value,
logger=self.logger
)
self._log("Adjusting boundaries...")
adjusted_map = adjust_boundary.adjust()
self._log("Adjusting boundaries... done")
self._log("Adjusting boundaries: succeeded")
return (True, adjusted_map)
