def extract_reid_features(reid_model, image, tlbrs):
if len(tlbrs) == 0:
return torch.FloatTensor()
patches = extract_image_patches(image, tlbrs)
patches = np.asarray(
[im_preprocess(cv2.resize(p, reid_model.inp_size)) for p in patches],
dtype=np.float32)
gpu = net_utils.get_device(reid_model)
with torch.no_grad():
im_var = Variable(torch.from_numpy(patches))
if gpu is not None:
im_var = im_var.cuda(gpu)
features = reid_model(im_var).data
return features
def extract_reid_features(reid_model, image, tlbrs):
if len(tlbrs) == 0:
return torch.FloatTensor()
patches = extract_image_patches(image, tlbrs)
gpu = net_utils.get_device(reid_model)
if LooseVersion(torch.__version__) > LooseVersion('0.3.1'):
with torch.no_grad():
im_var = Variable(torch.from_numpy(patches))
if gpu is not None:
im_var = im_var.cuda(gpu).float()
features = reid_model(im_var).data
else:
im_var = Variable(torch.from_numpy(patches), volatile=True)
if gpu is not None:
im_var = im_var.cuda(gpu)
features = reid_model(im_var).data
return features
def appearance_features(appearance_model, image, tlbrs):
if len(tlbrs) == 0:
return torch.FloatTensor()
candidates = extract_candidates(image, tlbrs)
candidates = np.asarray(
[im_preprocess(cv2.resize(c, (224, 224))) for c in candidates],
dtype=np.float32)
gpu = net_utils.get_device(appearance_model)
if LooseVersion(torch.__version__) > LooseVersion('0.3.1'):
with torch.no_grad():
im_var = Variable(torch.from_numpy(candidates))
if gpu is not None:
im_var = im_var.cuda(gpu)
features = appearance_model(im_var).data
else:
im_var = Variable(torch.from_numpy(candidates), volatile=True)
if gpu is not None:
im_var = im_var.cuda(gpu)
features = appearance_model(im_var).data
return features
