def test(**kwargs):
opt.parse(kwargs)
train_L, query_L, retrieval_L, train_x, query_x, retrieval_x, train_y, query_y, retrieval_y = load_data(
opt.data_path)
y_dim = query_y.shape[1]
print('...loading and splitting data finish')
img_model = ImgModule(opt.bit)
txt_model = TxtModule(y_dim, opt.bit)
if opt.load_img_path:
img_model.load(opt.load_img_path)
if opt.load_txt_path:
txt_model.load(opt.load_txt_path)
if opt.use_gpu:
img_model = img_model.cuda()
txt_model = txt_model.cuda()
qBX = generate_image_code(img_model, query_x, opt.bit)
qBY = generate_text_code(txt_model, query_y, opt.bit)
rBX = generate_image_code(img_model, retrieval_x, opt.bit)
rBY = generate_text_code(txt_model, retrieval_y, opt.bit)
if opt.use_gpu:
query_L = query_L.cuda()
retrieval_L = retrieval_L.cuda()
mapi2t = calc_map_k(qBX, rBY, query_L, retrieval_L)
mapt2i = calc_map_k(qBY, rBX, query_L, retrieval_L)
print('...test MAP: MAP(i->t): %3.3f, MAP(t->i): %3.3f' % (mapi2t, mapt2i))
def test(**kwargs):
opt.parse(kwargs)
images, tags, labels = load_data(opt.data_path)
y_dim = tags.shape[1]
X, Y, L = split_data(images, tags, labels)
print('...loading and splitting data finish')
img_model = ImgModule(opt.bit)
txt_model = TxtModule(y_dim, opt.bit)
if opt.load_img_path:
img_model.load(opt.load_img_path)
if opt.load_txt_path:
txt_model.load(opt.load_txt_path)
if opt.use_gpu:
img_model = img_model.cuda()
txt_model = txt_model.cuda()
print('-----------------------')
query_L = torch.from_numpy(L['query'])
query_x = torch.from_numpy(X['query'])
query_y = torch.from_numpy(Y['query'])
retrieval_L = torch.from_numpy(L['retrieval'])
retrieval_x = torch.from_numpy(X['retrieval'])
retrieval_y = torch.from_numpy(Y['retrieval'])
qBX = generate_image_code(img_model, query_x, opt.bit)
qBY = generate_text_code(txt_model, query_y, opt.bit)
rBX = generate_image_code(img_model, retrieval_x, opt.bit)
rBY = generate_text_code(txt_model, retrieval_y, opt.bit)
if opt.use_gpu:
query_L = query_L.cuda()
retrieval_L = retrieval_L.cuda()
mapi2t = calc_map_k(qBX, rBY, query_L, retrieval_L)
mapt2i = calc_map_k(qBY, rBX, query_L, retrieval_L)
print('...test MAP: MAP(i->t): %3.3f, MAP(t->i): %3.3f' % (mapi2t, mapt2i))
def debug(**kwargs):
opt.parse(kwargs)
# load data
images, tags, labels = load_data(opt.data_path)
y_dim = tags.shape[1]
labels1 = np.load(opt.l1_path)
X, Y, L, L1 = split_data(images, tags, labels, opt, labels1)
print('...loading and splitting data finish')
# init module
img_model = ImgModule(opt.bit)
txt_model = TxtModule(y_dim, opt.bit)
if opt.use_gpu:
img_model = img_model.cuda()
txt_model = txt_model.cuda()
print("load trained model from file..")
img_model.load(opt.load_img_path, use_gpu=True)
txt_model.load(opt.load_txt_path, use_gpu=True)
train_L = torch.from_numpy(L['train'])
train_L1 = torch.from_numpy(L1['train'])
train_x = torch.from_numpy(X['train'])
train_y = torch.from_numpy(Y['train'])
query_L = torch.from_numpy(L['query'])
query_x = torch.from_numpy(X['query'])
query_y = torch.from_numpy(Y['query'])
retrieval_L = torch.from_numpy(L['retrieval'])
retrieval_x = torch.from_numpy(X['retrieval'])
retrieval_y = torch.from_numpy(Y['retrieval'])
mapi2t, mapt2i = evaluate(img_model, txt_model, query_x, query_y, retrieval_x, retrieval_y, query_L, retrieval_L, opt.bit)
print('...test map: map(i->t): %3.3f, map(t->i): %3.3f' % (mapi2t, mapt2i))