The problem ScanX solves

ScanX addresses a growing concern in both education and professional settings: distinguishing AI-generated content from authentic human work. With the rise of AI tools like ChatGPT and advanced image and video generators, it's increasingly difficult to verify whether written assignments, images, or videos are genuine. This presents challenges in maintaining academic integrity and authenticity in media. ScanX leverages cutting-edge machine learning models to accurately detect whether text, images, or videos are AI-generated or real, ensuring that both digital and written submissions are fair and trustworthy.

Challenges I ran into

Since the models used are quite large, I pushed the backend on GitHub and uploaded the models separately on a drive link. I tried to use the aws hosting services but its free tier is only for CPU, not for GPU. Since our project requires GPU, we had to host the backend locally on GitHub.

It took a lot of efforts to make the page responsive and it was a long process in finally coming up with the final page designs.