The AI-based photo editor is an application that enhances your photos with the help of artificial intelligence. The application features several powerful tools to help you achieve professional-looking results with ease. for now, it was in implementation anyone who has interest in building this beautiful AI application can contribute to it.
Our AI algorithms are trained to autocorrect the color, denoise the image, add filters, and automatically adjust the brightness of your photos. With a simple and intuitive interface, you can easily apply these features with a single click.
Different libraries and tools used to build this project:
-
OpenCV: A computer vision library for image and video processing, used for reading and manipulating image data.
-
NumPy: A library for numerical operations, including array processing, used for transforming image data into arrays for processing.
-
TensorFlow or PyTorch: Machine learning frameworks for training and deploying deep learning models.
-
SciPy: A library for scientific computing, including image processing, used for image filtering and denoising.
-
Pillow: A library for image processing, used for reading, modifying, and saving image data.
-
Matplotlib: A plotting library for visualizing image data, used for debugging and validation.
-
Keras: A high-level neural networks API, built on top of TensorFlow, used for defining and training deep learning models.
-
Sklearn: A machine learning library for data analysis and modeling, used for preprocessing image data.
-
Seaborn: A data visualization library based on Matplotlib, used for generating and visualizing statistical models.
-
Scikit-image: An image processing library, used for image filtering, denoising, and feature extraction.
-
Flask: for deplying the entire project on webpage.
-
HTML & CSS: For creating the frontend part where users interect with the model.
CI/CCd tools like jenkins were also in testing stage for smooth deployment and productionizing the application.
To brief out the project, it was divided into 5 major sections which are as follows:
a. Automatic colour correction: adjusting the brightness, contrast, and colour balance of an image to make it more visually pleasing.
b. Image restoration: removing noise, blur, or other distortions from an image to improve its quality.
c. Object removal: removing unwanted objects from an image, such as a person or a piece of trash.
d. Image style transfer: applying the artistic style of one image to another image, such as making a photo look like a painting.
e. Face and object recognition: Identifying and tagging faces or objects in an image, such as identifying the different breeds of dogs in a picture
The first ipynb file shows the hierarchy of implementation, that ends with object identification and Tagging.