machine learning

Car Make and Model Classification System

• Developed a computer vision system utilizing deep learning to identify vehicle makes and models from images

• Leveraged transfer learning with ResNet50, InceptionV3, and EfficientNetB0 architectures on Stanford Cars Dataset

• Created model evaluation tools including accuracy comparisons, confusion matrices, and Grad-CAM visualizations

• Built a real-time classification interface with webcam integration for instant predictions

• Technologies: TensorFlow, OpenCV, Tkinter, Transfer Learning, CNN, Grad-CAM