MCastroFernandez/DeepLearningIntroduction

A gentle introduction using Pytorch (Work in Progress)

A gentle introduction to Deep Learning using PyTorch

... (Work In Progress)

Prerequisites

• Anaconda
• Git

Dependencies

1. PyTorch
   • version 1.9 or above
2. Torchvision
   • version 0.10 or above
3. Numpy
4. Matplotlib

Doc

Segmentation

• U-Net: https://arxiv.org/abs/1505.04597

Object detection:

• YOLO9000: https://arxiv.org/abs/1612.08242
• YOLOv3: https://arxiv.org/abs/1804.02767v1
• Object detection Review: https://arxiv.org/abs/1807.05511

Natural Language Processing:

• Recurrent Neural Networks:
  • TODO
• Transformers:
  • Original paper: https://arxiv.org/pdf/1706.03762.pdf
  • Positional Encoding: https://kazemnejad.com/blog/transformer_architecture_positional_encoding

Other:

• Backpropagation: https://mattmazur.com/2015/03/17/a-step-by-step-backpropagation-example/
• Distill: https://distill.pub/
• Convolutional arithmetic: https://arxiv.org/abs/1603.07285
• Loss functions: https://pytorch.org/docs/stable/nn.html#loss-functions
• Pytorch tutorials: https://pytorch.org/tutorials/