Inspiration As a gamer, I’ve always enjoyed spotting glitches, secrets, or unexpected differences in levels. That sharp attention to detail inspired me to build SpotTheGlitch—a tool that treats every image inspection like a quest to spot the odd one out and claim the win.
What it does SpotTheGlitch uses AI to scan images for subtle visual changes, errors, or inconsistencies. It’s designed to make quality checks quick, accurate, and even a little fun, by turning the process into a hunt for “glitches” just like you’d spot in your favorite games.
How we built it We used Python and PyTorch to handle image processing and model training. OpenCV helped with early difference detection, while PIL took care of prepping and cleaning up images. Iterative training and testing made sure the model learned to spot actual defects without being distracted by noise.
Challenges we ran into The biggest challenge was tuning the model to ignore random distractions in images and focus only on real issues. Sometimes it flagged harmless changes, like lighting or background movement, so we spent time refining the logic and cleaning the dataset.
Accomplishments that we're proud of We’re proud of creating a workflow that can detect subtle anomalies that even the human eye might miss. Plus, making the process feel like a game—quick, fresh, and rewarding—is exactly what we set out to do.
What we learned Along the way, we learned a ton about model tuning, image preprocessing, and how creativity from gaming can influence tech projects. We also got better at debugging (just like grinding through tough levels).
What's next for SpotTheGlitch Next up, we want to add a user-friendly dashboard and gamify the inspection workflow even further—think leaderboards, achievements for spotting rare glitches, and expanding into video analysis.
Built With
- flask-(for-web-interface)
- githu
- matplotlib
- numpy
- opencv
- pil-(pillow)
- python
- pytorch
- scikit-learn
