def display_whitening(self,save_string=None,save=True,normalize_A=True,max_factors=256,zerophasewhiten=True):
output = {}
nfilt, nchannels, szy, szx = self.convwhitenfiltershp
if zerophasewhiten:
A = self.zerophasewhitenmatrix
if hasattr(A,'get_value'):
A = A.get_value()
Afilter = self.convwhitenfilter.reshape((nfilt*nchannels,szy*szx))
Adefilter = self.convdewhitenfilter.reshape((nfilt*nchannels,szy*szx))
A = np.vstack((Afilter,Adefilter,A))
else:
A = self.whitenmatrix
if hasattr(A,'get_value'):
A = A.get_value()
A = A.T # make the columns the filters
psz = int(np.ceil(np.sqrt(A.shape[0])))
if not psz**2 == A.shape[0]:
A = np.vstack((A,np.zeros((psz**2 - A.shape[0],A.shape[1]))))
# plot the vectors in A
NN = min(A.shape[1],max_factors)
buf = 1
sz = int(np.sqrt(NN))
hval = np.max(np.abs(A))
array = -np.ones(((psz+buf)*sz+buf,(psz+buf)*sz+buf))
Aind = 0
for r in range(sz):
for c in range(sz):
if normalize_A:
hval = np.max(np.abs(A[:,Aind]))
Avalues = A[:,Aind].reshape(psz,psz)/hval
array[buf+(psz+buf)*c:buf+(psz+buf)*c+psz,buf+(psz+buf)*r:buf+(psz+buf)*r+psz] = Avalues
Aind += 1
hval = 1.
plt.figure(1)
plt.clf()
plt.imshow(array,vmin=-hval,vmax=hval,interpolation='nearest',cmap=plt.cm.gray)
plt.colorbar()
output['whitenmatrix'] = array
plt.draw()
if save:
from hdl.config import state_dir, tstring
savepath = os.path.join(state_dir,self.model_name + '_' + self.tstring)
if not os.path.isdir(savepath): os.makedirs(savepath)
if save_string is None:
save_string = tstring()
plt.figure(1)
fname = os.path.join(savepath, 'whitenmatrix_' + save_string + '.png')
plt.savefig(fname)
return output
def display(self,save_string=None,save=True,normalize_A=True,max_factors=256,zerophasewhiten=True):
output = {}
if hasattr(self.A,'get_value'):
A = self.A.get_value()
else:
A = self.A
if not zerophasewhiten:
Areshaped = np.reshape(A.T,self.kshp)
Aoutput = self._scipy_convolve4d(Areshaped,self.convdewhitenfilter,mode='same')
A = np.reshape(Aoutput,(self.kshp[0],np.prod(self.kshp[1:]))).T
psz = int(np.ceil(np.sqrt(A.shape[0])))
if not psz**2 == A.shape[0]:
A = np.vstack((A,np.zeros((psz**2 - A.shape[0],A.shape[1]))))
# plot the vectors in A
NN = min(self.NN,max_factors)
buf = 1
sz = int(np.sqrt(NN))
hval = np.max(np.abs(A))
array = -np.ones(((psz+buf)*sz+buf,(psz+buf)*sz+buf))
Aind = 0
for r in range(sz):
for c in range(sz):
if normalize_A:
hval = np.max(np.abs(A[:,Aind]))
Avalues = A[:,Aind].reshape(psz,psz)/hval
array[buf+(psz+buf)*c:buf+(psz+buf)*c+psz,buf+(psz+buf)*r:buf+(psz+buf)*r+psz] = Avalues
Aind += 1
hval = 1.
plt.figure(1)
plt.clf()
plt.imshow(array,vmin=-hval,vmax=hval,interpolation='nearest',cmap=plt.cm.gray)
plt.colorbar()
output['A'] = array
plt.draw()
if save:
from hdl.config import state_dir, tstring
savepath = os.path.join(state_dir,self.model_name + '_' + self.tstring)
if not os.path.isdir(savepath): os.makedirs(savepath)
if save_string is None:
save_string = tstring()
plt.figure(1)
fname = os.path.join(savepath, 'A_' + save_string + '.png')
plt.savefig(fname)
return output
def test_3Dvideo():
from matplotlib import pyplot as plt
test_name = '3Dvideo'
psz = 16
l = SGD(model=SparseSlowModel(patch_sz=psz,D=6*psz*psz,N=400,T=48),datasource='3Dvideo_color',batchsize=48)
batchsize = 48
databatch = l.get_databatch(batchsize)
batchsize = 1000
databatch = l.get_databatch(batchsize)
from hdl.config import tests_dir, tstring
savepath = os.path.join(tests_dir,test_name,tstring())
if not os.path.isdir(savepath): os.makedirs(savepath)
vidind = np.random.randint(l.video_buffer)
video = l.videos[vidind]
for tt in range(video.shape[0]):
plt.figure(1)
plt.clf()
plt.subplot(1,2,1)
frame = np.uint8(video[tt,:3,:,:].T)
plt.imshow(frame,interpolation='nearest')
plt.subplot(1,2,2)
frame = np.uint8(video[tt,3:,:,:].T)
plt.imshow(frame,interpolation='nearest')
fname = os.path.join(savepath, 'vid_' + str(vidind) + '_frame_%04d'%tt + '.png')
plt.savefig(fname)
batchsize = 48
databatch = l.get_databatch(batchsize)
for tt in range(batchsize):
plt.figure(1)
plt.clf()
plt.subplot(1,2,1)
frame_both = np.uint8(databatch[...,tt].reshape(6,l.model.patch_sz,l.model.patch_sz)).T
frame = frame_both[...,:3]
plt.imshow(frame,interpolation='nearest')
plt.subplot(1,2,2)
frame = frame_both[...,3:]
plt.imshow(frame,interpolation='nearest')
fname = os.path.join(savepath, 'databatch_frame_%04d'%tt + '.png')
plt.savefig(fname)
# test multiple loads
for tt in range(1000):
if not tt%100: print tt,
databatch = l.get_databatch(batchsize)
def test_YouTubeFaces():
from matplotlib import pyplot as plt
test_name = 'YouTubeFaces'
l = SGD(model=SparseSlowModel(patch_sz=48,N=400,T=64),datasource='YouTubeFaces_aligned',batchsize=64)
batchsize = 64
databatch = l.get_databatch(batchsize)
batchsize = 1000
databatch = l.get_databatch(batchsize)
from hdl.config import tests_dir, tstring
savepath = os.path.join(tests_dir,test_name,tstring())
if not os.path.isdir(savepath): os.makedirs(savepath)
vidind = int(np.floor(np.random.rand()*l.YouTubeInfo['num_videos']))
video = l.YouTubeInfo['videos'][vidind]
hval = np.abs(video).max()
for tt in range(video.shape[2]):
plt.figure(1)
plt.clf()
plt.imshow(video[:,:,tt],cmap=plt.cm.gray,vmin=-hval,vmax=hval,interpolation='nearest')
fname = os.path.join(savepath, 'vid_' + str(vidind) + '_frame_%04d'%tt + '.png')
plt.savefig(fname)
batchsize = 64
databatch = l.get_databatch(batchsize)
hval = np.abs(databatch).max()
for tt in range(batchsize):
plt.figure(1)
plt.clf()
plt.imshow(databatch[...,tt].reshape(l.model.patch_sz,l.model.patch_sz),cmap=plt.cm.gray,vmin=-hval,vmax=hval,interpolation='nearest')
fname = os.path.join(savepath, 'databatch_frame_%04d'%tt + '.png')
plt.savefig(fname)
# test multiple loads
for tt in range(100):
databatch = l.get_databatch(batchsize)
def test_vid075():
from matplotlib import pyplot as plt
test_name = 'vid075'
l = SGD(model=SparseSlowModel(patch_sz=20,N=400,T=64),datasource='vid075-chunks',batchsize=64)
batchsize = 64
databatch = l.get_databatch(batchsize)
batchsize = 1000
databatch = l.get_databatch(batchsize)
from hdl.config import tests_dir, tstring
savepath = os.path.join(tests_dir,test_name,tstring())
if not os.path.isdir(savepath): os.makedirs(savepath)
vidind = np.floor(np.random.rand()*l.nvideos)
hval = np.abs(l.videos[vidind,...]).max()
for tt in range(l.videot):
plt.figure(1)
plt.clf()
plt.imshow(l.videos[vidind,:,:,tt],cmap=plt.cm.gray,vmin=-hval,vmax=hval,interpolation='nearest')
fname = os.path.join(savepath, 'vid_' + str(vidind) + '_frame_'+ str(tt) + '.png')
plt.savefig(fname)
batchsize = 64
databatch = l.get_databatch(batchsize)
hval = np.abs(databatch).max()
for tt in range(batchsize):
plt.figure(1)
plt.clf()
plt.imshow(databatch[...,tt].reshape(l.model.patch_sz,l.model.patch_sz),cmap=plt.cm.gray,vmin=-hval,vmax=hval,interpolation='nearest')
fname = os.path.join(savepath, 'databatch_' + str(vidind) + '_frame_'+ str(tt) + '.png')
plt.savefig(fname)
import os
import hdl
reload(hdl)
from hdl.models import SparseSlowModel
from hdl.hdl import HDL
from hdl.config import tstring, state_dir
timestring = tstring()
model_base_name = 'HDL_loga_' + timestring + '/layer_%s'
m1 = SparseSlowModel()
layer1_name = 'SparseSlowModel_patchsz020_N512_NN512_l2_subspacel1_dist_2012-05-24_17-33-06/model.model'
fname = os.path.join(state_dir,layer1_name)
m1.load(fname)
m1.model_name = model_base_name % '1'
model_sequence = [
m1,
SparseSlowModel(patch_sz=None, N=512, T=48, sparse_cost='subspacel1', slow_cost='dist', perc_var=99.9, tstring=timestring, model_name=model_base_name % '2'),
SparseSlowModel(patch_sz=None, N=256, T=48, sparse_cost='l1', slow_cost=None, perc_var=99.9, tstring=timestring, model_name=model_base_name % '3')]
hdl_learner = HDL(model_sequence=model_sequence,datasource='PLoS09_Cars_Planes',output_function='proj_loga')
hdl_learner.learn(layer_start=1)
