- Project number
- Project / 011
- Year
- 2025
- Category
- AI Systems
- Status
- Completed
Plant Disease Classification
Transfer learning with CNNs to classify plant diseases from leaf images.
- Stack
- Python · TensorFlow · CNN
- Tags
- Deep Learning · Image Classification
01 /Overview
Compare building an image classifier from scratch with transfer learning, and measure the difference properly rather than from a single run.
02 /What I built
- Custom CNN: three convolution blocks with batch normalisation and a dense head
- VGG16 pre-trained on ImageNet, frozen, with a new classification head
- Five separate training phases per model for reliable averages
- Training and validation accuracy and loss recorded for every phase
03 /Outcome
VGG16 transfer learning reached 68.11% average validation accuracy against 48.07% for a custom CNN, across five training runs each.
04 /What I learned
Both models overfit (about 99% training accuracy), which made the case for regularisation and more varied data as clearly as the headline comparison.
