def mutual_info_metric_cars3d(vae, shapes_dataset):
dataset_loader = DataLoader(shapes_dataset,
batch_size=1000,
num_workers=1,
shuffle=False)
N = len(dataset_loader.dataset) # number of data samples
K = vae.z_dim # number of latent variables
nparams = vae.q_dist.nparams
vae.eval()
print('Computing q(z|x) distributions.')
qz_params = torch.Tensor(N, K, nparams)
n = 0
with torch.no_grad():
for xs in dataset_loader:
batch_size = xs.size(0)
xs = Variable(xs.view(batch_size, 3, 64, 64).cuda())
qz_params[n:n + batch_size] = vae.encoder.forward(xs).view(
batch_size, vae.z_dim, nparams).data
if vae.var_clipping:
qz_params.data[n:n + batch_size, ..., 1] = torch.clamp(
qz_params.data[n:n + batch_size, ..., 1],
vae.lowerbound_ln_var, vae.upperbound_ln_var)
n += batch_size
qz_params = Variable(qz_params.view(183, 4, 24, K, nparams).cuda())
qz_samples = vae.q_dist.sample(params=qz_params)
print('Estimating marginal entropies.')
# marginal entropies
marginal_entropies = estimate_entropies(
qz_samples.view(N, K).transpose(0, 1), qz_params.view(N, K, nparams),
vae.q_dist)
marginal_entropies = marginal_entropies.cpu()
cond_entropies = torch.zeros(3, K)
print('Estimating conditional entropies for yaw.')
for i in range(4):
qz_samples_pose_az = qz_samples[:, i, :, :].contiguous()
qz_params_pose_az = qz_params[:, i, :, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_pose_az.view(N // 4, K).transpose(0, 1),
qz_params_pose_az.view(N // 4, K, nparams), vae.q_dist)
cond_entropies[0] += cond_entropies_i.cpu() / 4
print('Estimating conditional entropies for pitch.')
for i in range(24):
qz_samples_pose_el = qz_samples[:, :, i, :].contiguous()
qz_params_pose_el = qz_params[:, :, i, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_pose_el.view(N // 24, K).transpose(0, 1),
qz_params_pose_el.view(N // 24, K, nparams), vae.q_dist)
cond_entropies[1] += cond_entropies_i.cpu() / 24
metric = compute_metric_faces(marginal_entropies, cond_entropies)
return metric, marginal_entropies, cond_entropies
def mutual_info_metric_faces(vae, shapes_dataset):
dataset_loader = DataLoader(shapes_dataset,
batch_size=1000,
num_workers=1,
shuffle=False)
N = len(dataset_loader.dataset) # number of data samples
K = vae.z_dim # number of latent variables
nparams = vae.q_dist.nparams
vae.eval()
print('Computing q(z|x) distributions.')
qz_params = torch.Tensor(N, K, nparams)
n = 0
for xs in dataset_loader:
batch_size = xs.size(0)
xs = Variable(xs.view(batch_size, 1, 64, 64), volatile=True)
qz_params[n:n + batch_size] = vae.encoder.forward(xs).view(
batch_size, vae.z_dim, nparams).data
n += batch_size
qz_params = Variable(qz_params.view(50, 21, 11, 11, K, nparams))
qz_samples = vae.q_dist.sample(params=qz_params)
print('Estimating marginal entropies.')
# marginal entropies
marginal_entropies = estimate_entropies(
qz_samples.view(N, K).transpose(0, 1), qz_params.view(N, K, nparams),
vae.q_dist)
marginal_entropies = marginal_entropies.cpu()
cond_entropies = torch.zeros(3, K)
print('Estimating conditional entropies for azimuth.')
for i in range(21):
qz_samples_pose_az = qz_samples[:, i, :, :, :].contiguous()
qz_params_pose_az = qz_params[:, i, :, :, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_pose_az.view(N // 21, K).transpose(0, 1),
qz_params_pose_az.view(N // 21, K, nparams), vae.q_dist)
cond_entropies[0] += cond_entropies_i.cpu() / 21
print('Estimating conditional entropies for elevation.')
for i in range(11):
qz_samples_pose_el = qz_samples[:, :, i, :, :].contiguous()
qz_params_pose_el = qz_params[:, :, i, :, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_pose_el.view(N // 11, K).transpose(0, 1),
qz_params_pose_el.view(N // 11, K, nparams), vae.q_dist)
cond_entropies[1] += cond_entropies_i.cpu() / 11
print('Estimating conditional entropies for lighting.')
for i in range(11):
qz_samples_lighting = qz_samples[:, :, :, i, :].contiguous()
qz_params_lighting = qz_params[:, :, :, i, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_lighting.view(N // 11, K).transpose(0, 1),
qz_params_lighting.view(N // 11, K, nparams), vae.q_dist)
cond_entropies[2] += cond_entropies_i.cpu() / 11
metric = compute_metric_faces(marginal_entropies, cond_entropies)
return metric, marginal_entropies, cond_entropies
def mutual_info_metric_faces(vae, shapes_dataset):
dataset_loader = DataLoader(shapes_dataset, batch_size=1000, num_workers=1, shuffle=False)
N = len(dataset_loader.dataset) # number of data samples
K = 10 # number of latent variables
nparams = dist.Normal().nparams
vae.eval()
print('Computing q(z|x) distributions.')
qz_params = torch.Tensor(N, K, nparams)
n = 0
for xs in dataset_loader:
batch_size = xs.size(0)
xs = xs.view(batch_size, 1, 64, 64).cuda()
z, mu, logvar, y = vae(xs)
mu = mu.view(batch_size, K, 1)
logvar = logvar.view(batch_size, K, 1)
target = torch.cat([mu, logvar], dim=2)
qz_params[n:n + batch_size] = target.view(batch_size, K, nparams).data
n += batch_size
qz_params = qz_params.view(50, 21, 11, 11, K, nparams).cuda()
qz_samples = dist.Normal().sample(params=qz_params)
print('Estimating marginal entropies.')
# marginal entropies
marginal_entropies = estimate_entropies(
qz_samples.view(N, K).transpose(0, 1),
qz_params.view(N, K, nparams),
dist.Normal())
marginal_entropies = marginal_entropies.cpu()
cond_entropies = torch.zeros(3, K)
print('Estimating conditional entropies for azimuth.')
for i in range(21):
qz_samples_pose_az = qz_samples[:, i, :, :, :].contiguous()
qz_params_pose_az = qz_params[:, i, :, :, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_pose_az.view(N // 21, K).transpose(0, 1),
qz_params_pose_az.view(N // 21, K, nparams),
dist.Normal())
cond_entropies[0] += cond_entropies_i.cpu() / 21
print('Estimating conditional entropies for elevation.')
for i in range(11):
qz_samples_pose_el = qz_samples[:, :, i, :, :].contiguous()
qz_params_pose_el = qz_params[:, :, i, :, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_pose_el.view(N // 11, K).transpose(0, 1),
qz_params_pose_el.view(N // 11, K, nparams),
dist.Normal())
cond_entropies[1] += cond_entropies_i.cpu() / 11
print('Estimating conditional entropies for lighting.')
for i in range(11):
qz_samples_lighting = qz_samples[:, :, :, i, :].contiguous()
qz_params_lighting = qz_params[:, :, :, i, :].contiguous()
cond_entropies_i = estimate_entropies(
qz_samples_lighting.view(N // 11, K).transpose(0, 1),
qz_params_lighting.view(N // 11, K, nparams),
dist.Normal())
cond_entropies[2] += cond_entropies_i.cpu() / 11
metric = compute_metric_faces(marginal_entropies, cond_entropies)
return metric, marginal_entropies, cond_entropies