Update app.py

app.py

@@ -1,7 +1,3 @@
-'''
-pip install extcolors
-'''
-
 import os
 import tensorflow as tf
 os.environ['TFHUB_MODEL_LOAD_FORMAT'] = 'COMPRESSED'
@@ -10,7 +6,6 @@ import PIL.Image
 import gradio as gr
 import tensorflow_hub as hub
 import matplotlib.pyplot as plt
-#from real_esrgan_app import *
 
 import gradio as gr
 import requests
@@ -34,14 +29,6 @@ import pathlib
 
 API_TOKEN = os.environ.get("HF_READ_TOKEN")
 
-'''
-dataset = load_dataset("Gustavosta/Stable-Diffusion-Prompts")
-prompt_df = dataset["train"].to_pandas()
-prompt_df = pd.read_csv("Stable-Diffusion-Prompts.csv")
-'''
-
-#DEFAULT_MODEL = "stabilityai/stable-diffusion-2-1"
-#DEFAULT_PROMPT = "1girl, aqua eyes, baseball cap, blonde hair, closed mouth, earrings, green background, hat, hoop earrings, jewelry, looking at viewer, shirt, short hair, simple background, solo, upper body, yellow shirt"
 DEFAULT_PROMPT = "X go to Istanbul"
 DEFAULT_ROLE = "Superman"
 DEFAULT_BOOK_COVER = "book_cover_dir/Blank.jpg"
@@ -60,67 +47,27 @@ def tensor_to_image(tensor):
 def perform_neural_transfer(content_image_input, style_image_input, hub_module = hub_module):
     content_image = content_image_input.astype(np.float32)[np.newaxis, ...] / 255.
     content_image = tf.image.resize(content_image, (400, 600))
-
-    #style_image_input = style_urls[style_image_input]
-    #style_image_input = plt.imread(style_image_input)
     style_image = style_image_input.astype(np.float32)[np.newaxis, ...] / 255.
-
     style_image = tf.image.resize(style_image, (256, 256))
-
     outputs = hub_module(tf.constant(content_image), tf.constant(style_image))
     stylized_image = outputs[0]
-
     stylized_image = tensor_to_image(stylized_image)
     content_image_input = tensor_to_image(content_image_input)
     stylized_image = stylized_image.resize(content_image_input.size)
     return stylized_image
-    '''
-    print("super_resolution_type :")
-    print(super_resolution_type)
-    #print(super_resolution_type.value)
-
-    if super_resolution_type not in ["base", "anime"]:
-        return stylized_image
-    else:
-        print("call else :")
-        stylized_image = inference(stylized_image, super_resolution_type)
-        return stylized_image
-    '''
 
 list_models = [
-    #"SDXL-1.0",
     "Pixel-Art-XL",
     "SD-1.5",
     "OpenJourney-V4",
     "Anything-V4",
     "Disney-Pixar-Cartoon",
     "Dalle-3-XL",
-    #"Midjourney-V4-XL",
 ]
 
-#list_prompts = get_samples()
 
 def generate_txt2img(current_model, prompt, is_negative=False, image_style="None style", steps=50, cfg_scale=7,
                      seed=None, API_TOKEN = API_TOKEN):
-
-    '''
-    if current_model == "SD-1.5":
-        API_URL = "https://api-inference.huggingface.co/models/runwayml/stable-diffusion-v1-5"
-    elif current_model == "SDXL-1.0":
-        API_URL = "https://api-inference.huggingface.co/models/stabilityai/stable-diffusion-xl-base-1.0"
-    elif current_model == "OpenJourney-V4":
-        API_URL = "https://api-inference.huggingface.co/models/prompthero/openjourney"
-    elif current_model == "Anything-V4":
-        API_URL = "https://api-inference.huggingface.co/models/xyn-ai/anything-v4.0"
-    elif current_model == "Disney-Pixar-Cartoon":
-        API_URL = "https://api-inference.huggingface.co/models/stablediffusionapi/disney-pixar-cartoon"
-    elif current_model == "Pixel-Art-XL":
-        API_URL = "https://api-inference.huggingface.co/models/nerijs/pixel-art-xl"
-    elif current_model == "Dalle-3-XL":
-        API_URL = "https://api-inference.huggingface.co/models/openskyml/dalle-3-xl"
-    elif current_model == "Midjourney-V4-XL":
-        API_URL = "https://api-inference.huggingface.co/models/openskyml/midjourney-v4-xl"
-    '''
     if current_model == "SD-1.5":
         API_URL = "https://api-inference.huggingface.co/models/runwayml/stable-diffusion-v1-5"
     elif current_model == "OpenJourney-V4":
@@ -324,20 +271,6 @@ def add_caption_on_image(input_image, caption, marg_ratio = 0.15, row_token_num
     marg_size = int(marg_ratio * max_image_size)
     colors, pixel_count = extcolors.extract_from_image(input_image)
     input_image = add_margin(input_image, marg_size, 0, 0, marg_size, colors[0][0])
-    '''
-    tmp_name = "{}.png".format(uuid1())
-    input_image.save(tmp_name)
-    ImageCaptioner.add_captions(tmp_name,
-                            caption,
-                            overwrite = 1,
-                            size = int(marg_size / 4),
-                            align = "TOP_LEFT",
-                            output = tmp_name,
-                            color = "black",
-                           )
-    output_image = Image.open(tmp_name)
-    os.remove(tmp_name)
-    '''
     font = ImageFont.truetype("DejaVuSerif-Italic.ttf" ,int(marg_size / 4))
     caption_token_list = list(map(lambda x: x.strip() ,caption.split(" ")))
     caption_list = list(map(" ".join ,batch_as_list(caption_token_list, row_token_num)))
@@ -349,7 +282,6 @@ def add_caption_on_image(input_image, caption, marg_ratio = 0.15, row_token_num
             (line_num + 1) * 1.1
         )))
         draw.text(position, line_caption, fill="black", font = font)
-
     return input_image
 
 
@@ -404,26 +336,6 @@ def make_video_from_image_list(image_list, video_name = "ppt.avi"):
     generate_video(image_list, video_name = video_name)
     return video_name
 
-'''
-style_transfer_client = Client("https://svjack-super-resolution-neural-style-transfer.hf.space")
-def style_transfer_func(content_img, style_img, style_transfer_client = style_transfer_client):
-    from uuid import uuid1
-    assert hasattr(content_img, "save")
-    assert hasattr(style_img, "save")
-    content_im_name = "{}.png".format(uuid1())
-    style_im_name = "{}.png".format(uuid1())
-    content_img.save(content_im_name)
-    style_img.save(style_im_name)
-    out = style_transfer_client.predict(
-    content_im_name,
-    style_im_name,
-    "none",
-    fn_index=1
-    )
-    os.remove(content_im_name)
-    os.remove(style_im_name)
-    return Image.open(out)
-'''
 def style_transfer_func(content_img, style_img):
     assert hasattr(content_img, "save")
     assert hasattr(style_img, "save")
@@ -469,10 +381,7 @@ def gen_images_from_event_fact(current_model, event_fact = DEFAULT_PROMPT, role_
 
 def image_click(images, evt: gr.SelectData,
     ):
-    #print(images)
-    #print(evt.index)
     img_selected = images[evt.index][0]["name"]
-    #print(img_selected)
     return img_selected
 
 def get_book_covers():
@@ -496,23 +405,16 @@ with gr.Blocks(css=".caption-label {display:none}") as demo:
         with gr.Column(elem_id="prompt-container"):
             current_model = gr.Dropdown(label="Current Model", choices=list_models, value="Pixel-Art-XL")
             style_reference_input_gallery = gr.Gallery(get_book_covers(),
-                            #width = 512,
-                            height = 512,
-                            label = "StoryBook Cover (click to use)")
+                            height = 768 + 64 + 32,
+                            label = "StoryBook Cover (click to use)",
+                            object_fit = "contain"
+                            )
         with gr.Column(elem_id="prompt-container"):
-            #with gr.Row(elem_id="prompt-container"):
             style_reference_input_image = gr.Image(
                             label = "StoryBook Cover (you can upload yourself or click from left gallery)",
-                            #width = 512,
                             value = DEFAULT_BOOK_COVER,
                             interactive = True,
                             )
-            '''
-            super_resolution_type = gr.Radio(choices = ["SD(Standard Definition)" ,"HD(High Definition)"],
-                        value="SD(Standard Definition)", label="Story Video Quality",
-                                                 interactive = True)
-            '''
-
             with gr.Row():
                 text_prompt = gr.Textbox(label="Event Prompt", placeholder=DEFAULT_PROMPT,
                     lines=1, elem_id="prompt-text-input", value = DEFAULT_PROMPT,
@@ -522,14 +424,12 @@ with gr.Blocks(css=".caption-label {display:none}") as demo:
                     elem_id="prompt-text-input", value = DEFAULT_ROLE,
                     info = "You should set the Role (X) with some famous man (like: Confucius Superman)"
                     )
+            with gr.Row():
                 text_button = gr.Button("Generate", variant='primary', elem_id="gen-button")
 
             with gr.Row():
-                #image_output = gr.Image(type="pil", label="Output Image", elem_id="gallery")
-                #image_output = gr.Gallery(label="Output Images", elem_id="gallery")
-                video_output = gr.Video(label = "Story Video", elem_id="gallery")
+                video_output = gr.Video(label = "Story Video", elem_id="gallery", height = 512,)
 
-    #text_button.click(generate_txt2img, inputs=[current_model, text_prompt, negative_prompt, image_style], outputs=image_output)
     style_reference_input_gallery.select(
             image_click, style_reference_input_gallery, style_reference_input_image
     )
@@ -537,7 +437,5 @@ with gr.Blocks(css=".caption-label {display:none}") as demo:
     text_button.click(gen_images_from_event_fact, inputs=[current_model, text_prompt, role_name, style_reference_input_image],
         outputs=video_output)
 
-    #select_button.click(generate_txt2img, inputs=[current_model, select_prompt, negative_prompt, image_style], outputs=image_output)
-    #demo.load(get_params, None, select_prompt)
 
 demo.launch(show_api=False)