Update app.py
Browse files
app.py
CHANGED
|
@@ -10,7 +10,13 @@ import time
|
|
| 10 |
|
| 11 |
# افزودن مسیر مورد نیاز برای ماژولهای CelebAMask-HQ
|
| 12 |
celebamask_path = "/home/user/app/CelebAMask-HQ"
|
| 13 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
# ایمپورت ماژولهای مورد نیاز
|
| 16 |
try:
|
|
@@ -21,6 +27,19 @@ try:
|
|
| 21 |
except ImportError as e:
|
| 22 |
IMPORT_SUCCESS = False
|
| 23 |
print(f"❌ Failed to import CelebAMask-HQ modules: {e}")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
# تنظیمات دستگاه
|
| 26 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
|
@@ -33,7 +52,7 @@ os.environ["HF_HOME"] = "/home/user/app/hf_cache"
|
|
| 33 |
transform = T.Compose([
|
| 34 |
T.Resize((512, 512), interpolation=PIL.Image.NEAREST),
|
| 35 |
T.ToTensor(),
|
| 36 |
-
T.Normalize((0.
|
| 37 |
])
|
| 38 |
|
| 39 |
# کلاسهای Face Parsing
|
|
@@ -63,8 +82,16 @@ class FaceParsingModel:
|
|
| 63 |
state_dict = torch.load(model_path, map_location="cpu")
|
| 64 |
|
| 65 |
# ایجاد مدل
|
| 66 |
-
self.model = unet()
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
self.model.eval()
|
| 69 |
self.model.to(self.device)
|
| 70 |
|
|
@@ -72,6 +99,8 @@ class FaceParsingModel:
|
|
| 72 |
|
| 73 |
except Exception as e:
|
| 74 |
print(f"❌ Failed to load model: {e}")
|
|
|
|
|
|
|
| 75 |
self.model = None
|
| 76 |
|
| 77 |
def predict(self, image):
|
|
@@ -96,15 +125,33 @@ class FaceParsingModel:
|
|
| 96 |
with torch.no_grad():
|
| 97 |
out = self.model(data)
|
| 98 |
out = generate_label(out, 512)
|
| 99 |
-
out = out[0].cpu().numpy()
|
| 100 |
-
|
| 101 |
-
|
|
|
|
|
|
|
| 102 |
# ترکیب تصویر اصلی با ماسک
|
| 103 |
-
resized_image = np.asarray(original_image.resize((512, 512)))
|
| 104 |
-
blended = resized_image * 0.
|
| 105 |
-
blended = np.clip(
|
| 106 |
-
|
| 107 |
-
return
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 108 |
|
| 109 |
def initialize_app():
|
| 110 |
"""Initialize application"""
|
|
@@ -112,12 +159,14 @@ def initialize_app():
|
|
| 112 |
|
| 113 |
print("[Info] PYTHONPATH:", os.environ.get("PYTHONPATH"))
|
| 114 |
print("[Info] CelebAMask-HQ path exists:", os.path.exists(celebamask_path))
|
| 115 |
-
print("[Info] face_parsing folder exists:", os.path.exists(
|
| 116 |
print("[Info] Module import success:", IMPORT_SUCCESS)
|
| 117 |
|
| 118 |
try:
|
| 119 |
face_parser = FaceParsingModel()
|
| 120 |
-
|
|
|
|
|
|
|
| 121 |
except Exception as e:
|
| 122 |
print(f"[Error] Initialization failed: {e}")
|
| 123 |
return False, f"Initialization failed: {e}", None
|
|
@@ -143,7 +192,6 @@ def process_image(input_image):
|
|
| 143 |
img_size = original_img.size
|
| 144 |
else:
|
| 145 |
img_size = input_image.size if hasattr(input_image, 'size') else input_image.shape[:2][::-1]
|
| 146 |
-
original_img = PIL.Image.fromarray(input_image) if isinstance(input_image, np.ndarray) else input_image
|
| 147 |
|
| 148 |
info_text = f"""
|
| 149 |
✅ پردازش انجام شد!
|
|
@@ -158,97 +206,8 @@ def process_image(input_image):
|
|
| 158 |
except Exception as e:
|
| 159 |
error_msg = f"❌ خطا در پردازش تصویر: {str(e)}"
|
| 160 |
print(error_msg)
|
|
|
|
|
|
|
| 161 |
return None, None, error_msg
|
| 162 |
|
| 163 |
-
|
| 164 |
-
"""ایجاد لیجند برای کلاسها"""
|
| 165 |
-
import matplotlib.pyplot as plt
|
| 166 |
-
|
| 167 |
-
legend_html = """
|
| 168 |
-
<div style='max-height: 300px; overflow-y: auto; border: 1px solid #ccc; padding: 10px; border-radius: 5px;'>
|
| 169 |
-
<h4>🎨 Legend - کلاسهای Face Parsing:</h4>
|
| 170 |
-
"""
|
| 171 |
-
|
| 172 |
-
colors = plt.get_cmap('tab20', len(CELEBA_CLASSES))
|
| 173 |
-
|
| 174 |
-
for i, class_name in enumerate(CELEBA_CLASSES):
|
| 175 |
-
color = colors(i)
|
| 176 |
-
color_hex = '#%02x%02x%02x' % (int(color[0]*255), int(color[1]*255), int(color[2]*255))
|
| 177 |
-
text_color = 'white' if color[0] * 0.299 + color[1] * 0.587 + color[2] * 0.114 < 0.5 else 'black'
|
| 178 |
-
legend_html += f"""
|
| 179 |
-
<div style='margin: 2px; padding: 5px; background-color: {color_hex}; color: {text_color}; border-radius: 3px;'>
|
| 180 |
-
<strong>{i}:</strong> {class_name}
|
| 181 |
-
</div>
|
| 182 |
-
"""
|
| 183 |
-
|
| 184 |
-
legend_html += "</div>"
|
| 185 |
-
return legend_html
|
| 186 |
-
|
| 187 |
-
# ایجاد اینترفیس Gradio
|
| 188 |
-
with gr.Blocks(title="CelebAMask-HQ Face Parsing", theme=gr.themes.Soft()) as demo:
|
| 189 |
-
gr.Markdown("""
|
| 190 |
-
# 🎭 CelebAMask-HQ Face Parsing Demo
|
| 191 |
-
**آپلود یک تصویر صورت و دریافت خروجی Face Parsing**
|
| 192 |
-
|
| 193 |
-
این مدل صورت را به 19 بخش مختلف تقسیم میکند (پوست، چشم، ابرو، بینی، دهان، مو و ...)
|
| 194 |
-
""")
|
| 195 |
-
|
| 196 |
-
with gr.Row():
|
| 197 |
-
with gr.Column():
|
| 198 |
-
input_image = gr.Image(
|
| 199 |
-
label="📷 تصویر ورودی",
|
| 200 |
-
type="filepath",
|
| 201 |
-
sources=["upload"],
|
| 202 |
-
height=300
|
| 203 |
-
)
|
| 204 |
-
process_btn = gr.Button("🚀 پردازش تصویر", variant="primary", size="lg")
|
| 205 |
-
|
| 206 |
-
with gr.Accordion("ℹ️ وضعیت برنامه", open=False):
|
| 207 |
-
status_display = gr.Markdown(f"""
|
| 208 |
-
**وضعیت:**
|
| 209 |
-
- 🎯 مدل: {'✅ لود شده' if success else '❌ خطا در لود'}
|
| 210 |
-
- 💻 دستگاه: `{device}`
|
| 211 |
-
- 📦 ماژولها: {'✅ ایمپورت شده' if IMPORT_SUCCESS else '❌ خطا در ایمپورت'}
|
| 212 |
-
- 🗂️ کلاسها: {len(CELEBA_CLASSES)}
|
| 213 |
-
""")
|
| 214 |
-
|
| 215 |
-
with gr.Column():
|
| 216 |
-
output_blended = gr.Image(
|
| 217 |
-
label="🎨 نتیجه ترکیبی (تصویر + ماسک)",
|
| 218 |
-
height=300
|
| 219 |
-
)
|
| 220 |
-
output_mask = gr.Image(
|
| 221 |
-
label="🎭 ماسک سگمنتیشن",
|
| 222 |
-
height=300
|
| 223 |
-
)
|
| 224 |
-
|
| 225 |
-
with gr.Row():
|
| 226 |
-
info_output = gr.Textbox(
|
| 227 |
-
label="📊 اطلاعات پردازش",
|
| 228 |
-
lines=3,
|
| 229 |
-
max_lines=6
|
| 230 |
-
)
|
| 231 |
-
|
| 232 |
-
with gr.Row():
|
| 233 |
-
gr.HTML(create_legend())
|
| 234 |
-
|
| 235 |
-
# اتصال رویدادها
|
| 236 |
-
process_btn.click(
|
| 237 |
-
fn=process_image,
|
| 238 |
-
inputs=[input_image],
|
| 239 |
-
outputs=[output_blended, output_mask, info_output]
|
| 240 |
-
)
|
| 241 |
-
|
| 242 |
-
input_image.upload(
|
| 243 |
-
fn=process_image,
|
| 244 |
-
inputs=[input_image],
|
| 245 |
-
outputs=[output_blended, output_mask, info_output]
|
| 246 |
-
)
|
| 247 |
-
|
| 248 |
-
if __name__ == "__main__":
|
| 249 |
-
print("🚀 Starting Face Parsing Application...")
|
| 250 |
-
demo.launch(
|
| 251 |
-
server_name="0.0.0.0",
|
| 252 |
-
server_port=7860,
|
| 253 |
-
share=False
|
| 254 |
-
)
|
|
|
|
| 10 |
|
| 11 |
# افزودن مسیر مورد نیاز برای ماژولهای CelebAMask-HQ
|
| 12 |
celebamask_path = "/home/user/app/CelebAMask-HQ"
|
| 13 |
+
face_parsing_path = os.path.join(celebamask_path, "face_parsing")
|
| 14 |
+
sys.path.insert(0, celebamask_path)
|
| 15 |
+
sys.path.insert(0, face_parsing_path)
|
| 16 |
+
|
| 17 |
+
print("Python path:", sys.path)
|
| 18 |
+
print("CelebAMask path exists:", os.path.exists(celebamask_path))
|
| 19 |
+
print("Face parsing path exists:", os.path.exists(face_parsing_path))
|
| 20 |
|
| 21 |
# ایمپورت ماژولهای مورد نیاز
|
| 22 |
try:
|
|
|
|
| 27 |
except ImportError as e:
|
| 28 |
IMPORT_SUCCESS = False
|
| 29 |
print(f"❌ Failed to import CelebAMask-HQ modules: {e}")
|
| 30 |
+
# تعریف توابع جایگزین در صورت نیاز
|
| 31 |
+
def unet(**kwargs):
|
| 32 |
+
from unet import UNet
|
| 33 |
+
return UNet(**kwargs)
|
| 34 |
+
|
| 35 |
+
def generate_label(inputs, imsize=512):
|
| 36 |
+
pred_batch = []
|
| 37 |
+
for input in inputs:
|
| 38 |
+
pred = np.squeeze(input.data.max(1)[1].cpu().numpy(), axis=0)
|
| 39 |
+
pred_batch.append(pred)
|
| 40 |
+
pred_batch = np.array(pred_batch)
|
| 41 |
+
pred_batch = torch.from_numpy(pred_batch)
|
| 42 |
+
return pred_batch
|
| 43 |
|
| 44 |
# تنظیمات دستگاه
|
| 45 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
|
|
|
| 52 |
transform = T.Compose([
|
| 53 |
T.Resize((512, 512), interpolation=PIL.Image.NEAREST),
|
| 54 |
T.ToTensor(),
|
| 55 |
+
T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)),
|
| 56 |
])
|
| 57 |
|
| 58 |
# کلاسهای Face Parsing
|
|
|
|
| 82 |
state_dict = torch.load(model_path, map_location="cpu")
|
| 83 |
|
| 84 |
# ایجاد مدل
|
| 85 |
+
self.model = unet(n_channels=3, n_classes=19)
|
| 86 |
+
|
| 87 |
+
# تطبیق state dict در صورت نیاز
|
| 88 |
+
new_state_dict = {}
|
| 89 |
+
for k, v in state_dict.items():
|
| 90 |
+
if k.startswith('module.'):
|
| 91 |
+
k = k[7:] # حذف 'module.' اگر وجود دارد
|
| 92 |
+
new_state_dict[k] = v
|
| 93 |
+
|
| 94 |
+
self.model.load_state_dict(new_state_dict, strict=False)
|
| 95 |
self.model.eval()
|
| 96 |
self.model.to(self.device)
|
| 97 |
|
|
|
|
| 99 |
|
| 100 |
except Exception as e:
|
| 101 |
print(f"❌ Failed to load model: {e}")
|
| 102 |
+
import traceback
|
| 103 |
+
traceback.print_exc()
|
| 104 |
self.model = None
|
| 105 |
|
| 106 |
def predict(self, image):
|
|
|
|
| 125 |
with torch.no_grad():
|
| 126 |
out = self.model(data)
|
| 127 |
out = generate_label(out, 512)
|
| 128 |
+
out = out[0].cpu().numpy()
|
| 129 |
+
|
| 130 |
+
# تبدیل به تصویر رنگی
|
| 131 |
+
colored_mask = self.colorize_mask(out)
|
| 132 |
+
|
| 133 |
# ترکیب تصویر اصلی با ماسک
|
| 134 |
+
resized_image = np.asarray(original_image.resize((512, 512)))
|
| 135 |
+
blended = resized_image * 0.6 + colored_mask * 0.4
|
| 136 |
+
blended = np.clip(blended, 0, 255).astype(np.uint8)
|
| 137 |
+
|
| 138 |
+
return colored_mask, blended
|
| 139 |
+
|
| 140 |
+
def colorize_mask(self, mask):
|
| 141 |
+
"""رنگآمیزی ماسک بر اساس کلاسها"""
|
| 142 |
+
# پالت رنگ برای 19 کلاس
|
| 143 |
+
palette = [
|
| 144 |
+
[0, 0, 0], [255, 0, 0], [0, 255, 0], [0, 0, 255], [255, 255, 0],
|
| 145 |
+
[255, 0, 255], [0, 255, 255], [128, 0, 0], [0, 128, 0], [0, 0, 128],
|
| 146 |
+
[128, 128, 0], [128, 0, 128], [0, 128, 128], [128, 128, 128], [255, 128, 0],
|
| 147 |
+
[255, 0, 128], [128, 255, 0], [0, 255, 128], [255, 128, 128]
|
| 148 |
+
]
|
| 149 |
+
|
| 150 |
+
colored = np.zeros((mask.shape[0], mask.shape[1], 3), dtype=np.uint8)
|
| 151 |
+
for i in range(len(palette)):
|
| 152 |
+
colored[mask == i] = palette[i]
|
| 153 |
+
|
| 154 |
+
return colored
|
| 155 |
|
| 156 |
def initialize_app():
|
| 157 |
"""Initialize application"""
|
|
|
|
| 159 |
|
| 160 |
print("[Info] PYTHONPATH:", os.environ.get("PYTHONPATH"))
|
| 161 |
print("[Info] CelebAMask-HQ path exists:", os.path.exists(celebamask_path))
|
| 162 |
+
print("[Info] face_parsing folder exists:", os.path.exists(face_parsing_path))
|
| 163 |
print("[Info] Module import success:", IMPORT_SUCCESS)
|
| 164 |
|
| 165 |
try:
|
| 166 |
face_parser = FaceParsingModel()
|
| 167 |
+
success = face_parser.model is not None
|
| 168 |
+
status_msg = "Model loaded successfully" if success else "Model failed to load"
|
| 169 |
+
return success, status_msg, face_parser
|
| 170 |
except Exception as e:
|
| 171 |
print(f"[Error] Initialization failed: {e}")
|
| 172 |
return False, f"Initialization failed: {e}", None
|
|
|
|
| 192 |
img_size = original_img.size
|
| 193 |
else:
|
| 194 |
img_size = input_image.size if hasattr(input_image, 'size') else input_image.shape[:2][::-1]
|
|
|
|
| 195 |
|
| 196 |
info_text = f"""
|
| 197 |
✅ پردازش انجام شد!
|
|
|
|
| 206 |
except Exception as e:
|
| 207 |
error_msg = f"❌ خطا در پردازش تصویر: {str(e)}"
|
| 208 |
print(error_msg)
|
| 209 |
+
import traceback
|
| 210 |
+
traceback.print_exc()
|
| 211 |
return None, None, error_msg
|
| 212 |
|
| 213 |
+
# ادامه کد مشابه قبل برای Gradio interface...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|