Generative AI for Visual Tasks

This repository contains a Jupyter Notebook (task-1.ipynb) that demonstrates the use of Generative AI and image classification techniques for various visual tasks. The notebook covers the following topics:

1. Generative AI for Visuals

In this section, the notebook showcases the use of Stable Diffusion, a state-of-the-art generative AI model, for generating images from text prompts. The code includes:

Installation of required libraries (accelerate, diffusers, and keras_cv)
Loading and configuring the Stable Diffusion model
A function to generate images from text prompts
Example usage of the image generation function with the prompt "Car"
Plotting the generated images using matplotlib

2. Product Recognition (Image Classification)

This section focuses on image classification using pre-trained models like EfficientNetV2B0 and MobileNetV2. The code includes:

Loading the pre-trained EfficientNetV2B0 model for image classification
Functions for preprocessing and classifying images
Example usage of the classification functions on the previously generated images
Commented code for using the MobileNetV2 model for classification

3. Exclusion of Non-Relevant Images

In this section, the notebook provides a mechanism to filter out images that do not contain any of the specified target products. This ensures that only relevant visuals are processed and enhanced. The code includes:

A function find_target_products that searches for the target products in the classification results
Example usage of the find_target_products function with a target product "sports_car"

Usage

To use this notebook, follow these steps:

Clone the repository to your local machine.
Install the required dependencies (accelerate, diffusers, keras_cv, tensorflow, numpy, pandas, and matplotlib).
Open the task-1.ipynb notebook in your preferred Jupyter environment.
Run the notebook cells to execute the code and explore the different sections.

Note: Due to the large size of the pre-trained models and computational requirements, some cells may take a significant amount of time to execute, particularly on systems with limited resources.

Acknowledgemet