← Back to portfolio
Image Classification with CNN
Custom CNN architecture for image classification, trained with data augmentation.
Problem
Manually reviewing and tagging large image datasets doesn't scale. An automated classifier needs to generalize beyond its training set, not just memorize it.
Approach
Designed a custom Convolutional Neural Network with convolutional, pooling, and dropout layers in TensorFlow/Keras, and trained it with data augmentation to improve generalization.
Results
- Achieved strong generalization on unseen data
Want the full picture, or think this kind of work fits a role you're hiring for?