CNNs

Materials:

Date: TBD

Pre-work:

1. Refresh ML foundations.
2. Read “The 100 page ML book” by Andiry Burkov. Chapters accessible here
3. FFNs

In-Class

1. CNNs Intro
   • Chapter 7[R05] covers topics like Invariance, Channels, Convolution Operation, Padding, Stride, Pooling.
   • CNNs[R02]
2. Visualizations
   • CNNs[R02]
3. CNN Architectures
   • Chapter 8[R04] covers landmark CNN architectures such as AlexNet, VGG, ResNet, DenseNet.

Lab

1. 1d CNNs (tbd)
2. 2d CNNs (tbd)

Post-class:

1. [notebook] ID Convolution R05
2. [notebook] ID Convolution for MNSIT R05
3. [notebook] 2D Convolution R05
4. [notebook] CNN for MNIST R05
5. [youtube] Lectures from R07
   • CNNs: An Intro Part-1, part-2
   • Backprop in CNNS youtube
   • Evolution of CNNs for Image Classification part-1, part-2
   • Recent Architectures youtube [as of 2021]
   • CNNs for Object Detection Part-1, Part-2 Part-3
   • CNNs for Segmentation youtube

Papers

• AlexNet - ImageNet Classification with Deep Convolutional Neural Networks
• ZFNet Visualizing and Understanding Convolutional Networks
• VGGNet Very Deep Convolutional Networks for Large-Scale Image Recognition
• GooLeNet Going Deeper with Convolutions
• ResNet Deep Residual Learning for Image Recognition
• RCNN Rich feature hierarchies for accurate object detection and semantic segmentation -Faster-RCNN Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• YOLO You Only Look Once: Unified, Real-Time Object Detection
• UNet U-Net: Convolutional Networks for Biomedical Image Segmentation