AmrDabour/image-colorizer

Image Colorization Model

This model colorizes grayscale images using a Stable Diffusion ControlNet approach.

Usage

# Install required dependencies
pip install diffusers tensorflow torch

# Load and use the model
from huggingface_hub import hf_hub_download
import tensorflow as tf
from tensorflow import keras
import os

# Define the ColorizeModel class (same as in the original notebook)
# [Insert the ColorizeModel class definition here]

# Download the model files
model_path = hf_hub_download(repo_id="AmrDabour/image-colorizer", filename="image_colorizer.h5")
component_path = os.path.dirname(os.path.dirname(model_path)) + "/models--AmrDabour--image-colorizer/snapshots/COMMIT_HASH"

# Load and use the model
with keras.utils.custom_object_scope({'ColorizeModel': ColorizeModel}):
    loaded_model = keras.models.load_model(model_path, compile=False)
    result = loaded_model.colorize("your_grayscale_image.jpg", output_path="colored_output.png")

Replace AmrDabour and COMMIT_HASH with appropriate values.