Notes

prompt = "((photo of dcnwtxhg)) Headshot portrait of a man in Business casual clothing with suit and tie, realistic face, professional attire, confident expression, natural lighting, sharp focus, high-resolution (4k), professional photography, clean and polished look, with a outdoor background"
negative_prompt = "sectarian mark, deformed iris, random other person, deformed pupils, semirealistic, sgi, 3d,dual man, render sketch, cartoon, drawing, anime:1.4), text, closeup, cropped, out of frame, worst quality, low quality, jpeg artifact, ugly, duplicate, morbid, mutilated, extra finger, blurry, extra limbs, cloned face, missing legs, extra arm, many finger, long neck"
num_samples = 4
guidance_scale = 7.5
num_inference_steps = 24
height = 512
width = 512


# Define the path to the folder where you want to save the images
output_folder = "/home/ubuntu/socialitecomm/output_tom"

# Set the initial count of generated images
generated_images = 0

verified_folder = "/home/ubuntu/socialitecomm/Good_images"

while len(os.listdir(verified_folder)) < 140:
with autocast("cuda"), torch.inference_mode():
images = pipe(
prompt,
height=height,
width=width,
negative_prompt=negative_prompt,
num_images_per_prompt=num_samples,
num_inference_steps=num_inference_steps,
guidance_scale=guidance_scale,
generator=g_cuda
).images

# Save the images to the output folder with random names
for i, img in enumerate(images):
# Generate a random filename
random_chars = ''.join(random.choices(string.ascii_letters + string.digits, k=10))
img_path = os.path.join(output_folder, f"output_{random_chars}.png")
img.save(img_path)
print(f"Image {i+1} saved successfully: {img_path}")




from deepface import DeepFace
import os
import shutil


# Load pre-trained models
models = "Facenet512"
DeepFace.build_model(models)

# Path to the folder containing the photos
folder_path = "/home/ubuntu/socialitecomm/output_tom"

for filename in os.listdir(folder_path):
if filename.endswith(".jpg") or filename.endswith(".png"):
# Full path to the photo
img_path = os.path.join(folder_path, filename)

try:
# Perform face detection
img_pathh = "/home/ubuntu/socialitecomm/greg_image/IMG_8722.jpg"
detections = DeepFace.detectFace(img_pathh, detector_backend="opencv")

# Check if face detection is successful
if len(detections) == 0:
# Remove the photo from the folder
os.remove(img_path)
print(f"Photo {filename} does not contain a detectable face. Removed from folder.")
continue

# Perform face recognition
result = DeepFace.verify("/home/ubuntu/socialitecomm/greg_image/IMG_8722.jpg", img_path, model_name="Facenet512", detector_backend = "opencv")

# Extract the matching percentage
matching_percentage = 1 - result["distance"]

# Update verified field based on matching percentage
result["verified"] = matching_percentage > 0.50

# Display the result
print(f"Matching Percentage: {matching_percentage*100:.2f}%, Verification result for {filename}: {result}")


import subprocess
command = "nvidia-smi"
result = subprocess.run(command, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
if result.returncode == 0:
print(result.stdout)
else:
print("Error:", result.stderr)

if result["verified"]:
# Move the verified image to a new folder
# verified_folder = "/content/drive/MyDrive/Good_images"
# shutil.move(img_path, os.path.join(verified_folder, filename))
verified_folder = "/home/ubuntu/socialitecomm/Good_images"
new_filename = f"image_{matching_percentage:.2f}_{filename}"
new_filepath = os.path.join(verified_folder, new_filename)
shutil.move(img_path, new_filepath)

# Rename the file
# os.rename(new_filepath, os.path.join(verified_folder, new_filename))
else:
# Move the unverified image to another folder
unverified_folder = "/home/ubuntu/socialitecomm/Bad_images"
new_filename = f"image_{matching_percentage:.2f}_{filename}"
new_filepath = os.path.join(unverified_folder, new_filename)
shutil.move(img_path, new_filepath)

except Exception as e:
print(f"Error processing photo {filename}: {str(e)}")
generated_images += len(images)