Certificate of Completion

What You Learned

• Explain how digital images are represented as pixel arrays and preprocessed for computer vision models
• Build CNN architectures that detect spatial patterns through convolutional filters, pooling, and feature hierarchies
• Compare object detection approaches from two-stage (Faster R-CNN) to one-stage (YOLO) and their speed-accuracy trade-offs
• Distinguish semantic, instance, and panoptic segmentation and identify when to use each approach
• Apply transfer learning and data augmentation to train accurate models with limited labeled images
• Evaluate computer vision applications across healthcare, manufacturing, and autonomous vehicles — including ethical risks