![]() Save_image(tf.squeeze(fake_image), filename="Super Resolution")Įvaluating Performance of the Model !wget "" -O test.jpg Plot_image(tf.squeeze(fake_image), title="Super Resolution") ![]() Print("Time Taken: %f" % (time.time() - start)) Save_image(tf.squeeze(hr_image), filename="Original Image") Plot_image(tf.squeeze(hr_image), title="Original Image") Performing Super Resolution of images loaded from path hr_image = preprocess_image(IMAGE_PATH) Image = omarray(tf.cast(image, tf.uint8).numpy()) Hr_image = tf.image.crop_to_bounding_box(hr_image, 0, 0, hr_size, hr_size) Hr_size = (tf.convert_to_tensor(hr_image.shape) // 4) * 4 Hr_image = tf.code_image(tf.io.read_file(image_path)) """ Loads image from path and preprocesses to make it model ready 12:10:50 (19.2 MB/s) - ‘original.png’ saved ĭefining Helper Functions def preprocess_image(image_path): Model trained on DIV2K Dataset (on bicubically downsampled images) on image patches of size 128 x 128. (Preferrably bicubically downsampled images). This colab demonstrates use of TensorFlow Hub Module for Enhanced Super Resolution Generative Adversarial Network ( by Xintao Wang et.al.) įor image enhancing.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |