Check basic implementations on CIFAR10 in the Deep Learning Lab project here
Goals:
- Implement some basic convolutional networks
- Implement different data augmentation
- Implement VGG model
-
Residual Nets example of Keras
-
Residual Nets example of Pytorch
-
Wide Resnet
-
Dense Nets
Images from "Labeled Faces in the Wild" dataset (LFW) in realistic scenarios, poses and gestures. Faces are automatically detected and cropped to 100x100 pixels RGB.
Training set: 10585 images
Test set: 2648 images
Python Notebook: here
Python code: here
Goals:
-
Implement a model with >95% accuracy over test set
-
Implement a model with >90% accuracy with less than 100K parameters
get some inspiration from Paper
Images of 20 different models of cars.
Training set: 791 images
Test set: 784 images
-
Version 1. Two different CNNs:
Python code: here
-
Version 2. The same CNN (potentially a pre-trained model)
Python code: here
Goals:
-
Understand the above Keras implementations:
- Name the layers
- Built several models
- Understand tensors sizes
- Connect models with operations (outproduct)
- Create an image generator that returns a list of tensors
- Create a data flow with multiple inputs for the model
- Understand the limitations of the proposed solution
-
Lab Project (2 points of labs mark)
- Load a pre-trained VGG16
- Connect this pre-trained model and form a bi-linear
- Train freezing weights or not
- ...
Code extracted and adapted from github
Goals:
-
Use a more simple version from: alpha version
-
Use the full version. Code adapted to download images for training and test:
Python Notebook: here
Python code: here
-
Understand the above Keras implementations:
- How to load the inception net
- How to merge encoder and inception result
- Use image functions to obtain lab space
- Create an appropiate data augmentation
Need help? Read
Code extracted and adapted from github
Content image
Style image
Result image
Python Notebook: here
Python code: here
You are welcome!