PixelForge

AI enhanced image editing on the go!

PixelForge is a cutting-edge prototype that enables on-the-go image editing powered by AI/ML models. It utilizes an ESP32 camera, which is remotely linked to Dropbox via an API, automatically storing each captured image. OpenShiftAI then retrieves the image data from Dropbox using an access token. From there, users can seamlessly apply one of three available stylization models to transform their images.

Role: Cloud Leader - Utilized the RedHat OpenShift Software to develop the cloud interface and AI model integration.

See my write-up on the OpenShift Platform here.

Technologies used:

  • ESP-32 CAM (Image Capture)
  • Dropbox API
  • RedHat OpenShift Cloud Platform
  • Intel AMX Hardware Acceleration
  • Fast Neural Style Transfer Models

Skills: Python Programming, C Programming, Cloud Infrastructure, TensorFlow, PyTorch

All the project code is fully open-source!

Preview of the Frontend after Deployment!

Video Demo: Checkout or sleep-deprived walkthrough of the system!