Overview
I successfully completed "Convolutional Neural Networks," the fourth course in the Deep Learning Specialization on Coursera. This course focused on building and training CNNs for visual recognition tasks and explored state-of-the-art architectures in computer vision.
The course emphasized:
- Foundations of CNNs: Understanding convolution, padding, strides, and pooling layers.
- Practical Implementation: Building CNNs using frameworks like TensorFlow.
- Modern Architectures: Insight into LeNet, AlexNet, VGG, ResNet, and Inception.
- Applications: Including face recognition, image classification, and neural style transfer.
Key Concepts Covered
- Convolutional and pooling layers
- Padding and stride mechanics
- CNN architecture design
- Residual Networks (ResNets)
- Object detection and localization
- Face verification and recognition systems
- Data augmentation
- Transfer learning
- 1x1 convolutions and Network-in-Network
- Neural Style Transfer