Technologies used:
Blood Cells Classification 🩸🔬 Link to heading
In this project, we successfully classified 96x96 RGB images of blood cells into eight distinct classes, each representing a particular cell state. This multi-class classification problem required us to assign the correct class label to each RGB image.
Martian Terrain Semantic Segmentation 🪐🔍 Link to heading
In this project, we developed an ensemble model for segmenting Martian terrain images. The ensemble combines multiple U-Net models with custom enhancements to improve segmentation accuracy.