def _test11():
# A reduced version of test10
ret = ""
# GmmMgr setup
num_states = 2
dimension = 2
models = []
for i in xrange(num_states):
dm = DummyModel(dimension, 1.0)
models.append(dm)
gmm_mgr = GmmMgr(models)
gb = GraphBuilder()
node_id0 = gb.new_node((0, 0))
node_id1 = gb.new_node((1, 1))
node_id2 = gb.new_node((2, 1))
node_id3 = gb.new_node((3, 1))
node_id4 = gb.new_node((4, 2))
# The topology here is slightly complex than the previous example
arc_id = gb.new_arc(node_id0, node_id1)
arc_id = gb.new_arc(node_id1, node_id4)
arc_id = gb.new_arc(node_id0, node_id2)
arc_id = gb.new_arc(node_id2, node_id3)
arc_id = gb.new_arc(node_id3, node_id4)
arc_id = gb.new_arc(node_id2, node_id4)
gr0 = FrozenGraph(gb)
# Make two Hmms with 3 states and order 2 (self loop, forward 1)
# The models in the middle are special and can skip.
seed(0)
hmm0 = make_forward_hmm(gmm_mgr, num_states, order=2, exact=False)
hmm1 = Hmm(1)
trans = array(((0.0, 0.5, 0.5), (0.0, 0.5, 0.5), (0.0, 0.0, 0.0)))
hmm1.build_model(gmm_mgr, (0, ), 1, 1, trans)
hmm2 = make_forward_hmm(gmm_mgr, num_states, order=2, exact=True)
hmm_mgr = HmmMgr((hmm0, hmm1, hmm2))
spd = {}
spd[(0, 1)] = (0.4, )
spd[(0, 2)] = (0.6, )
spd[(2, 3)] = (0.4, )
spd[(2, 4)] = (0.6, )
tg0 = TrainingGraph(gr0, hmm_mgr, split_prob_dict=spd)
if do_display:
tg0.dot_display()
tg0.dot_display(expand_hmms=True)
with DebugPrint("bwt_ctsh") if True else DebugPrint():
result_hmm = tg0.convert_to_standalone_hmm()
ret += "\n\n========= TG CONVERTED TO Hmm =========\n\n" + result_hmm.to_string(
full=True)
return ret
def on():
utt.on
utt.scale = 5
prints = (
#'gaussian',
#'gaussian_pt',
'SimpleGaussianTrainer',
'ddt_accum',
)
db=DebugPrint(logstream, *prints)
db.on()
audio.on
def _test9():
# Like test8, but now HMMs have multiple inputs and outputs.
ret = ""
# GmmMgr setup
num_states = 3
dimension = 2
models = []
for i in xrange(num_states):
dm = DummyModel(dimension, 1.0)
models.append(dm)
gmm_mgr = GmmMgr(models)
gb = GraphBuilder()
node_id0 = gb.new_node((0, 0))
node_id1 = gb.new_node((1, 1))
node_id2 = gb.new_node((2, 1))
node_id3 = gb.new_node((3, 1))
node_id4 = gb.new_node((4, 1))
node_id5 = gb.new_node((5, 2))
arc_id = gb.new_arc(node_id0, node_id1)
arc_id = gb.new_arc(node_id1, node_id2)
arc_id = gb.new_arc(node_id2, node_id3)
arc_id = gb.new_arc(node_id3, node_id4)
arc_id = gb.new_arc(node_id4, node_id5)
gr0 = FrozenGraph(gb)
# Make two Hmms with 3 states and order 3 (self loop, forward 1, forward 2)
# The models in the middle are special and can skip directly
seed(0)
hmm0 = make_forward_hmm(gmm_mgr, num_states, order=3, exact=True)
hmm1 = Hmm(1)
trans = array(((0.0, 0.0, 0.0, 0.5, 0.5, 0.0,
0.0), (0.0, 0.0, 0.0, 0.5, 0.0, 0.5,
0.0), (0.0, 0.0, 0.0, 0.5, 0.0, 0.0,
0.5), (0.0, 0.0, 0.0, 0.5, 0.35, 0.1, 0.05),
(0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
0.0), (0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
0.0), (0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)))
hmm1.build_model(gmm_mgr, (0, ), 3, 3, trans)
hmm2 = make_forward_hmm(gmm_mgr, num_states, order=3, exact=True)
hmm_mgr = HmmMgr((hmm0, hmm1, hmm2))
with DebugPrint("bwt_vrfy") if False else DebugPrint():
tg0 = TrainingGraph(gr0, hmm_mgr, split_prob_dict=dict())
result_hmm = tg0.convert_to_standalone_hmm()
ret += "\n\n========= TG CONVERTED TO Hmm =========\n\n" + result_hmm.to_string(
full=True)
return ret
def fifo_test():
from time import sleep
at = AccTracker(sending=False)
right = ('A', ((1, 'A'), (0, 'B')))
wrong = ('B', ((1, 'A'), (0, 'B')))
input = tuple(([right]* 25 + [wrong] * 10) * 100)
with open('/tmp/fifo.acc', 'w') as fifo:
with DebugPrint(fifo, 'acc_track'):
while 1:
sleep(0.1)
for item in input:
at.process(item)
class DpToggle(object):
"""
An object with two attributes, on and off, that toggle debug printing
of the DebugPrint object created from the constructor args.
"""
# XXX DebugPrint needs work to develop semantics to support multiple
# instances of DpToggle....
def __init__(self, *args):
self.dp_control = DebugPrint(*args)
self._on = False
def __del__(self):
self.off
@property
def on(self):
if not self._on:
self._on = True
self.dp_control.on()
return 'on'
@property
def off(self):
if self._on:
self._on = False
self.dp_control.off()
return 'off'
def startup(args):
global outfile
global on, off, dp
global start, stop
global processor
from functools import partial
from onyx.audio.liveaudio import inputs, default_input, LiveAudioSource
from onyx.signalprocessing.htkmfcc import make_fft_abs_processor
from onyx.util.debugprint import DebugPrint, dcheck, dprint
#from onyx.util.streamprocess import ChainProcessor
if '-o' in args:
outfile = open(args[args.index('-o')+1], 'wb', 0)
else:
outfile = sys.stdout
if '-l' in args:
logfile = open(args[args.index('-l')+1], 'wb', 0)
else:
logfile = sys.stdout
## if '-m' in args:
## mic_name = args[args.index('-m')+1]
## else:
## mic_name = 'blue'
debug = DebugPrint(logfile, 'saddemo')
## on = doer(debug.on)
## off = doer(debug.off)
dp = partial(dprint, 'saddemo', DebugPrint.NO_PREFIX, ' ')
plot_points = 255
line = [' '] * plot_points
## # select a live input
## mic_id = default_input()
## for id, chan, name, manu in inputs():
## if name.lower().find(mic_name.lower()) != -1:
## mic_id = id
## break
## mic = LiveAudioSource(mic_id, verbose=False)
## start = doer(mic.start)
## stop = doer(mic.stop)
return
# use the module's mic
mic = audio.mic
# XXX LiveAudioSource (and thus audiomodule.cpp) need to be able to provide
# the sample frequency....
sample_nsec = 22676 # 44100 Hz
#frame_nsec = 10000000 # 10 msec
#window_nsec = 25600000 # 25.6 msec
frame_nsec = 8000000 # 8 msec
window_nsec = 24000000 # 24 msec
dither = 1 / (1 << 10)
preemcoef = 0.96875
samples_per_sec, fft_size, fftmag = make_fft_abs_processor(sample_nsec, frame_nsec, window_nsec,
dither=dither, preemcoef=preemcoef,
zmeansource=True, usehamming=True,
max_channels=2)
print 'samples_per_sec', samples_per_sec, ' fft_size', fft_size
# figure out range of fft bins to work with
lowfreq = 300
highfreq = 4000
import math
low_index = int(math.ceil(lowfreq * fft_size / samples_per_sec))
high_index = int(math.floor(highfreq * fft_size / samples_per_sec))
# the bin selector
select = slice(low_index, high_index)
def display(data):
band = data[select]
band *= band
sum = float(band.sum())
dB = int(dB_scale * (10 * math.log10(sum) + dB_offset))
dp('%11.6f ' % sum, '%3d' % dB)
dB = max(1, dB)
#outfile.write('\n' + ' ' * dB + '|')
outfile.write( pen * dB + 'O' + '\n')
mic.set_sendee(fftmag.process)
fftmag.set_sendee(display)
def __init__(self, *args):
self.dp_control = DebugPrint(*args)
self._on = False
def _test10():
# Like test9, but now HMMs are arranged in a diamond pattern so inter-HMM
# probabilities come into play
ret = ""
# GmmMgr setup
num_states = 3
dimension = 2
models = []
for i in xrange(num_states):
dm = DummyModel(dimension, 1.0)
models.append(dm)
gmm_mgr = GmmMgr(models)
gb = GraphBuilder()
node_id0 = gb.new_node((0, 0))
node_id1 = gb.new_node((1, 1))
node_id2 = gb.new_node((2, 1))
node_id3 = gb.new_node((3, 1))
node_id4 = gb.new_node((4, 1))
node_id5 = gb.new_node((5, 2))
# The topology here is more complex than previous examples
arc_id = gb.new_arc(node_id0, node_id1)
arc_id = gb.new_arc(node_id1, node_id5)
arc_id = gb.new_arc(node_id0, node_id2)
arc_id = gb.new_arc(node_id2, node_id3)
arc_id = gb.new_arc(node_id3, node_id4)
arc_id = gb.new_arc(node_id3, node_id5)
arc_id = gb.new_arc(node_id4, node_id5)
gr0 = FrozenGraph(gb)
# Make two Hmms with 3 states and order 3 (self loop, forward 1, forward 2)
# The models in the middle are special and can skip.
seed(0)
hmm0 = make_forward_hmm(gmm_mgr, num_states, order=3, exact=True)
hmm1 = Hmm(1)
trans = array(((0.0, 0.0, 0.0, 0.5, 0.5, 0.0,
0.0), (0.0, 0.0, 0.0, 0.5, 0.0, 0.5,
0.0), (0.0, 0.0, 0.0, 0.5, 0.0, 0.0,
0.5), (0.0, 0.0, 0.0, 0.5, 0.35, 0.1, 0.05),
(0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
0.0), (0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
0.0), (0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0)))
hmm1.build_model(gmm_mgr, (0, ), 3, 3, trans)
hmm2 = make_forward_hmm(gmm_mgr, num_states, order=3, exact=True)
hmm_mgr = HmmMgr((hmm0, hmm1, hmm2))
spd = {}
spd[(0, 1)] = (0.4, 0.3, 0.8)
spd[(0, 2)] = (0.6, 0.7, 0.2)
spd[(3, 4)] = (0.4, 0.3, 0.8)
spd[(3, 5)] = (0.6, 0.7, 0.2)
tg0 = TrainingGraph(gr0, hmm_mgr, split_prob_dict=spd)
with DebugPrint("bwt_ctsh") if True else DebugPrint():
result_hmm = tg0.convert_to_standalone_hmm()
ret += "\n\n========= TG CONVERTED TO Hmm =========\n\n" + result_hmm.to_string(
full=True)
return ret
exit()
numpy.set_printoptions(linewidth=300)
## print "================================ TEST 0 ==========================="
## with DebugPrint("bwt_ctsh", "hmm_gxfs", "hmm_da", "hmm_pofb", "hmm_poff") if True else DebugPrint():
## s0 = test0(10)
## print s0
## print "================================ TEST 1 ==========================="
## with DebugPrint("bwt_ctsh") if True else DebugPrint():
## s1 = test1(10)
## print s1
## print "================================ TEST 2 ==========================="
## s2 = test2(10)
## print s2
## print "================================ TEST 3 ==========================="
## s3 = test3(10)
## print s3
## print "================================ TEST 4 ==========================="
## s4 = test4(4, 10)
## print s4
print "================================ TEST 5 ==========================="
with DebugPrint("bwt_ctsh", "hmm_gxfs", "hmm_da", "hmm_pofb",
"hmm_poff") if True else DebugPrint():
s5 = test5(1, do_display=True)
print s5
