def single_image_upload(): # Access the user-sent image from the request object. fname = request.files['input-file'].filename fpath = os.path.join("imgs", fname) request.files['input-file'].save(fpath) # Create a dataset to contain the image. ds = Dataset(name='dataset_' + str(int(random() * 1000000)), prefix='./imgs', batch_size=1, images=True) ds.initialize(sources=[fname]) ds_id = STATE.add_dataset(ds) return json.dumps({'datasetId': ds_id})
def test_keras_feature_extractor_extract_features(): ext = KerasFeatureExtractor(TEST_NET_ID, ckpt_path=TEST_CKPT_PATH) ds = Dataset(name=TEST_NAME, prefix=TEST_PREFIX, batch_size=8) ds.initialize(fp=TEST_SOURCES) ds.load_images() imgs = [e.image for e in ds.elements] prepro = Preprocessor() imgs = prepro.preprocess_images(imgs) result = ext.extract_features(images=imgs) assert isinstance(result, np.ndarray) == True assert len(result) == ds.count
def test_prepro(): from data import Dataset ds = Dataset(name='ds', prefix='/home/sam/Pictures', batch_size=1) ds.initialize(sources=['puppy.jpg']) p1 = Preprocessor(224, 224, PreproMode.RescaleHeightPadOrCropWidth) p2 = Preprocessor(224, 224, PreproMode.RescaleWidthPadOrCropHeight) p3 = Preprocessor(224, 224, PreproMode.RescaleWidthRescaleHeight) p4 = Preprocessor(224, 224, PreproMode.AspectRatioCrop) p5 = Preprocessor(224, 224, PreproMode.AspectRatioPad) i1 = p1.preprocess(ds)[0] i2 = p2.preprocess(ds)[0] i3 = p3.preprocess(ds)[0] i4 = p4.preprocess(ds)[0] i5 = p5.preprocess(ds)[0] i1.show() input() i2.show() input() i3.show() input() i4.show() input() i5.show() return
def make_initialized_test_dataset(): ds = Dataset(name=TEST_NAME, prefix=TEST_PREFIX, batch_size=8) ds.initialize(fp=TEST_SOURCES) return ds