Update app.py

app.py

@@ -3,168 +3,204 @@
 # import os
 # import random
 # from PIL import Image
+# import shutil
+# import requests
 
-# # Paths
-# OUTPUT_DIR = "outputs"
-# MODEL_PATH = "blazer_model.pkl"  # Adjusted for local Hugging Face repo structure
+# # === Setup Paths ===
+# BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+# GEN_SCRIPT = os.path.join(BASE_DIR, "stylegan3", "gen_images.py")
+# OUTPUT_DIR = os.path.join(BASE_DIR, "outputs")
+# MODEL_PATH = os.path.join(BASE_DIR, "blazer_model.pkl")
+# SAVE_DIR = os.path.join(BASE_DIR, "saved_images")
 
-# # Ensure the output directory exists
 # os.makedirs(OUTPUT_DIR, exist_ok=True)
+# os.makedirs(SAVE_DIR, exist_ok=True)
 
-# # Function to generate images using StyleGAN3
+# # === Image Generation Function ===
 # def generate_images():
-#     command = f"python stylegan3/gen_images.py --outdir={OUTPUT_DIR} --trunc=1 --seeds='3-5,7,9,12-14,16-26,29,31,32,34,40,41' --network={MODEL_PATH}"
+#     command = [
+#         "python",
+#         GEN_SCRIPT,
+#         f"--outdir={OUTPUT_DIR}",
+#         "--trunc=1",
+#         "--seeds=3-5,7,9,12-14,16-26,29,31,32,34,40,41",
+#         f"--network={MODEL_PATH}"
+#     ]
 #     try:
-#         subprocess.run(command, shell=True, check=True)
+#         subprocess.run(command, check=True, capture_output=True, text=True)
 #     except subprocess.CalledProcessError as e:
-#         return f"Error generating images: {e}"
+#         return f"Error generating images:\n{e.stderr}"
 
-# # Function to select 5 random images
+# # === Select Random Images from Output Folder ===
 # def get_random_images():
 #     image_files = [f for f in os.listdir(OUTPUT_DIR) if f.endswith(".png")]
 #     if len(image_files) < 10:
 #         generate_images()
 #         image_files = [f for f in os.listdir(OUTPUT_DIR) if f.endswith(".png")]
 #     random_images = random.sample(image_files, min(10, len(image_files)))
-#     return [Image.open(os.path.join(OUTPUT_DIR, img)) for img in random_images]
+#     image_paths = [os.path.join(OUTPUT_DIR, img) for img in random_images]
+#     return image_paths
 
-# # Gradio function
-# def generate_and_display():
-#     generate_images()
-#     return get_random_images()
+# # === Send Image to Backend ===
+# def send_to_backend(img_path, user_id):
+#     if not user_id:
+#         return "❌ user_id not found in URL."
 
-# # UI
+#     if not img_path or not os.path.exists(img_path):
+#         return "⚠️ No image selected or image not found."
+
+#     try:
+#         with open(img_path, 'rb') as f:
+#             files = {'file': ('generated_image.png', f, 'image/png')}
+
+#             # Your backend endpoint here
+#             url = f" https://7da2-2409-4042-6e81-1806-de6-b8e5-836c-6b95.ngrok-free.app/images/upload/{user_id}"
+#             response = requests.post(url, files=files)
+
+#         if response.status_code == 201:
+#             return "✅ Image uploaded and saved to database!"
+#         else:
+#             return f"❌ Upload failed: {response.status_code} - {response.text}"
+
+#     except Exception as e:
+#         return f"⚠️ Error: {str(e)}"
+
+# # === Gradio Interface ===
 # with gr.Blocks() as demo:
 #     gr.Markdown("# 🎨 AI-Generated Clothing Designs - Blazers")
+
 #     generate_button = gr.Button("Generate New Designs")
-#     output_gallery = gr.Gallery(label="Generated Designs", columns=5, rows=2)
-#     generate_button.click(fn=generate_and_display, outputs=output_gallery)
+#     user_id_state = gr.State()
+
+#     @demo.load(inputs=None, outputs=[user_id_state])
+#     def get_user_id(request: gr.Request):
+#         return request.query_params.get("user_id", "")
+
+#     image_components = []
+#     file_paths = []
+#     save_buttons = []
+#     outputs = []
+
+#     # Use 3 columns layout
+#     for row_idx in range(4):  # 4 rows (to cover 10 images)
+#         with gr.Row():
+#             for col_idx in range(3):  # 3 columns
+#                 i = row_idx * 3 + col_idx
+#                 if i >= 10:
+#                     break
+#                 with gr.Column():
+#                     img = gr.Image(width=180, height=180, label=f"Design {i+1}")
+#                     image_components.append(img)
+
+#                     file_path = gr.Textbox(visible=False)
+#                     file_paths.append(file_path)
+
+#                     save_btn = gr.Button("💾 Save to DB")
+#                     save_buttons.append(save_btn)
+
+#                     output = gr.Textbox(label="Status", interactive=False)
+#                     outputs.append(output)
+
+#                     save_btn.click(
+#                         fn=send_to_backend,
+#                         inputs=[file_path, user_id_state],
+#                         outputs=output
+#                     )
+
+#     # Generate button logicg
+#     def generate_and_display_images():
+#         image_paths = get_random_images()
+#         return image_paths + image_paths  # One for display, one for hidden path tracking
+
+#     generate_button.click(
+#         fn=generate_and_display_images,
+#         outputs=image_components + file_paths
+#     )
 
 # if __name__ == "__main__":
 #     demo.launch()
 
-
-
-
-import gradio as gr
-import subprocess
-import os
-import random
+import torch
+from transformers import CLIPModel, CLIPProcessor
 from PIL import Image
-import shutil
-import requests
-
-# === Setup Paths ===
-BASE_DIR = os.path.dirname(os.path.abspath(__file__))
-GEN_SCRIPT = os.path.join(BASE_DIR, "stylegan3", "gen_images.py")
-OUTPUT_DIR = os.path.join(BASE_DIR, "outputs")
-MODEL_PATH = os.path.join(BASE_DIR, "blazer_model.pkl")
-SAVE_DIR = os.path.join(BASE_DIR, "saved_images")
-
-os.makedirs(OUTPUT_DIR, exist_ok=True)
-os.makedirs(SAVE_DIR, exist_ok=True)
-
-# === Image Generation Function ===
-def generate_images():
-    command = [
-        "python",
-        GEN_SCRIPT,
-        f"--outdir={OUTPUT_DIR}",
-        "--trunc=1",
-        "--seeds=3-5,7,9,12-14,16-26,29,31,32,34,40,41",
-        f"--network={MODEL_PATH}"
-    ]
-    try:
-        subprocess.run(command, check=True, capture_output=True, text=True)
-    except subprocess.CalledProcessError as e:
-        return f"Error generating images:\n{e.stderr}"
-
-# === Select Random Images from Output Folder ===
-def get_random_images():
-    image_files = [f for f in os.listdir(OUTPUT_DIR) if f.endswith(".png")]
-    if len(image_files) < 10:
-        generate_images()
-        image_files = [f for f in os.listdir(OUTPUT_DIR) if f.endswith(".png")]
-    random_images = random.sample(image_files, min(10, len(image_files)))
-    image_paths = [os.path.join(OUTPUT_DIR, img) for img in random_images]
-    return image_paths
-
-# === Send Image to Backend ===
-def send_to_backend(img_path, user_id):
-    if not user_id:
-        return "❌ user_id not found in URL."
-
-    if not img_path or not os.path.exists(img_path):
-        return "⚠️ No image selected or image not found."
-
-    try:
-        with open(img_path, 'rb') as f:
-            files = {'file': ('generated_image.png', f, 'image/png')}
-
-            # Your backend endpoint here
-            url = f" https://7da2-2409-4042-6e81-1806-de6-b8e5-836c-6b95.ngrok-free.app/images/upload/{user_id}"
-            response = requests.post(url, files=files)
-
-        if response.status_code == 201:
-            return "✅ Image uploaded and saved to database!"
-        else:
-            return f"❌ Upload failed: {response.status_code} - {response.text}"
-
-    except Exception as e:
-        return f"⚠️ Error: {str(e)}"
-
-# === Gradio Interface ===
-with gr.Blocks() as demo:
-    gr.Markdown("# 🎨 AI-Generated Clothing Designs - Blazers")
-
-    generate_button = gr.Button("Generate New Designs")
-    user_id_state = gr.State()
-
-    @demo.load(inputs=None, outputs=[user_id_state])
-    def get_user_id(request: gr.Request):
-        return request.query_params.get("user_id", "")
-
-    image_components = []
-    file_paths = []
-    save_buttons = []
-    outputs = []
-
-    # Use 3 columns layout
-    for row_idx in range(4):  # 4 rows (to cover 10 images)
-        with gr.Row():
-            for col_idx in range(3):  # 3 columns
-                i = row_idx * 3 + col_idx
-                if i >= 10:
-                    break
-                with gr.Column():
-                    img = gr.Image(width=180, height=180, label=f"Design {i+1}")
-                    image_components.append(img)
-
-                    file_path = gr.Textbox(visible=False)
-                    file_paths.append(file_path)
-
-                    save_btn = gr.Button("💾 Save to DB")
-                    save_buttons.append(save_btn)
-
-                    output = gr.Textbox(label="Status", interactive=False)
-                    outputs.append(output)
-
-                    save_btn.click(
-                        fn=send_to_backend,
-                        inputs=[file_path, user_id_state],
-                        outputs=output
-                    )
-
-    # Generate button logicg
-    def generate_and_display_images():
-        image_paths = get_random_images()
-        return image_paths + image_paths  # One for display, one for hidden path tracking
-
-    generate_button.click(
-        fn=generate_and_display_images,
-        outputs=image_components + file_paths
+import numpy as np
+import pickle
+import gradio as gr
+
+# Force CPU usage for optimization
+device = torch.device("cpu")
+
+# Load your GAN model
+with open("/content/drive/MyDrive/clothGAN/blazer_model.pkl", "rb") as f:
+    G = pickle.load(f)['G_ema'].eval().cpu()  # Ensure model is in eval mode and on CPU
+
+# Load CLIP model and processor
+clip_model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32").eval().cpu()
+clip_processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
+
+# Generate images
+def generate_images(G, num_images=10):  # Reduce for CPU performance
+    z = torch.randn(num_images, G.z_dim)
+    c = None
+    with torch.no_grad():
+        images = G(z, c)
+        images = (images.clamp(-1, 1) + 1) * (255 / 2)
+        images = images.permute(0, 2, 3, 1).numpy().astype(np.uint8)
+    return z, images
+
+# Rank images using CLIP
+def rank_by_clip(images, prompt, top_k=3):  # Reduce top_k for speed
+    images_pil = [Image.fromarray(img) for img in images]
+    inputs = clip_processor(text=[prompt], images=images_pil, return_tensors="pt", padding=True)
+
+    with torch.no_grad():
+        image_features = clip_model.get_image_features(pixel_values=inputs["pixel_values"])
+        text_features = clip_model.get_text_features(input_ids=inputs["input_ids"], attention_mask=inputs["attention_mask"])
+
+        image_features = image_features / image_features.norm(dim=-1, keepdim=True)
+        text_features = text_features / text_features.norm(dim=-1, keepdim=True)
+
+        similarity = (image_features @ text_features.T).squeeze()
+
+    top_indices = similarity.argsort(descending=True)[:top_k]
+    best_images = [images_pil[i] for i in top_indices]
+    return best_images
+
+# Gradio interface function
+def generate_top_dresses(prompt):
+    _, images = generate_images(G, num_images=20)
+    top_images = rank_by_clip(images, prompt, top_k=2)
+    return top_images
+
+# Launch Gradio
+with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue")) as demo:
+    gr.Markdown("""
+    # 👔 AI Blazer Generator  
+    _Type in your dream outfit, and let the AI bring your fashion vision to life!_  
+    Just describe your ideal blazer, and see how AI transforms your words into fashion.
+    """)
+
+    with gr.Row():
+        input_box = gr.Textbox(
+            label="Describe your blazer",
+            placeholder="e.g., 'Beige blazer with gold buttons and slim fit'",
+            lines=2
+        )
+
+    with gr.Row():
+        submit_button = gr.Button("Generate Designs")
+
+    output_gallery = gr.Gallery(label="Generated Designs", show_label=True, columns=2, height="auto")
+
+
+    examples = gr.Examples(
+        examples = [
+    ["Black velvet blazer"],
+    ["Pink blazer with buttons"]
+],
+        inputs=[input_box]
     )
 
-if __name__ == "__main__":
-    demo.launch()
+    submit_button.click(fn=generate_top_dresses, inputs=input_box, outputs=output_gallery)
 
+demo.launch()