Improving My AI Image Enhancer App with Interactive Features

From a Simple Script to a More Dynamic Experience

As part of the MIT online course “Designing and Developing AI Products and Services”, I’ve been working on a project to explore how artificial intelligence can improve image quality. The app uses the Real-ESRGAN model to restore and enhance low-resolution or blurred photos.

Today, I decided to add new features to make it more interactive and visual. What started as a simple coding exercise to strengthen my design and technical skills is slowly turning into a practical tool that shows how AI can improve images, and how much you can learn by building.

New Features Added

To make the app easier to use and experiment with, I added two key improvements:

Interactive Parameters

Users can now enter parameters directly from the console instead of modifying the code. This makes it easy to test different combinations and immediately see how they affect the output.

With this feature, the model dynamically constructs the command for Real-ESRGAN, adapting to user inputs and allowing for quick iterations and fine-tuning.

Before and After Image Preview

To make the results more intuitive, I added a visualization step using PIL and Matplotlib. This displays the original image and the enhanced version side by side, making the improvements instantly visible.

This addition transforms the workflow from a text-based script into a visual and interactive experience, closer to how end users would interact with an actual AI product.

What I Learned

Adding these features helped me understand:

• How user interaction enhances the learning and experimentation process.
• The importance of visual feedback when demonstrating AI model performance.
• How small design choices can dramatically improve usability, even in a prototype.

These are the same principles taught in the MIT AI Product Design course — focusing on usability, feedback loops, and product thinking rather than just technical execution.

Next Steps

My next goal is to build a simple web interface for the app so users can upload images and test enhancements directly from their browser. This will make it easier to demonstrate how AI can be applied to real-world problems.

Every iteration of this project reinforces a simple idea: learning AI through creation. Small improvements, constant testing, and visual feedback, that’s how progress feels tangible.