BROWSE BY TECHNOLOGY

NumPy · scikit-learn · metrics & confusion matrix projects

Final-year project topics built with NumPy · scikit-learn · metrics & confusion matrix — working builds, abstracts and honest quotation pricing.

Street-view house number photo with extracted digit crops and the recognised number displayed alongside
Built to order

House Number Recognition using SVHN

This project builds a digit-recognition pipeline that reads house numbers from street photos. A CNN digit classifier trained...

Matplotlib · Seaborn · training curves NumPy · scikit-learn · metrics & confusion matrix Flask · demo web app with image upload
Request a quotation
Field photo of an insect pest on a leaf alongside the demo app interface showing the predicted pest identity with confidence scores.
Built to order

Insect Pest Identification using CNN

This project builds an image-classification system that identifies crop insect pests from field photographs. A transfer-learn...

Matplotlib · Seaborn · training curves Python 3.10 · PyTorch · transfer learning torchvision · PIL · OpenCV · image preprocessing
Request a quotation
Parking lot camera frame with slot polygons marked free and occupied by the detector
Built to order

Parking Slot Occupancy Detection using CNN

This project builds a vision pipeline that classifies parking slots as free or occupied from camera frames. A CNN binary clas...

Matplotlib · Seaborn · training curves NumPy · scikit-learn · metrics & confusion matrix PyTorch .pth weights · exported from training run
Request a quotation
Solar panel cell electroluminescence image with a CNN defect-detection overlay highlighting cracks and hotspot regions.
Built to order

Solar Panel Defect Detection using CNN

This project builds an inspection assistant that spots solar-cell defects from electroluminescence (EL) imagery. A ResNet or...

NumPy · scikit-learn · metrics & confusion matrix Python 3.10 · PyTorch · CNN training & inference torchvision · pretrained ResNet & EfficientNet backbones
Request a quotation
Demo interface showing a car photo with the predicted damage assessment and an estimated repair-cost panel.
Built to order

Vehicle Damage Assessment using CNN

This project builds a transfer-learning CNN that categorizes visible car damage — dent, scratch, cracked bumper, broken headl...

Matplotlib · Seaborn · training curves Python 3.10 · PyTorch · transfer learning torchvision · PIL · OpenCV · image preprocessing
Request a quotation

Have a different idea? Request a custom project