s46144603 VQVAE OASIS#478

Open

joshuamcallan wants to merge 21 commits intoshakes76:topic-recognitionfrom

joshuamcallan:topic-recognition

Commits on Oct 24, 2022

Loaded OASIS dataset from Cdrive and seperated into train, test and validate - each preprocessed by normalising by factoring by 255.
joshuamcallan
committed
Created a simple encoder for the VQVAE to reduce the amount of paramters in the model.
joshuamcallan
committed
Added a simple decoder network for the VQVAE.
joshuamcallan
committed

Commits on Oct 25, 2022

Added VectorQuantization class in order to create a discrete latent space.
joshuamcallan
committed
Added a vqvae model by feeding inputs into the encoder and the encoder through vq_layer for the quantised vae and finally fed the quantised vae through the decoder for the model.
joshuamcallan
committed
Added training loop for VQVAE in order to backpropagate and record losses.
joshuamcallan
committed
Added predicitions using the test set based on the VQVAE model.
joshuamcallan
committed
Put the training model into the wrong file (predict.py) so swapped it to the train.py file and also added plots of the losses and metrics during training.
joshuamcallan
committed
Removed the training model from this file and added functionality to compare original vs reconstructed images as well as calculate the structured similarity (SSIM) of these two images).
joshuamcallan
committed
Added a description of the VQVAE model and how it works. Also added a main title and description of the problem. Some headings were also made.
joshuamcallan
committed
Added more description for implementation of dataset as well as module dependencies and references.
joshuamcallan
committed
When choosing the random images, it was selecting from the train dataset so fixed this to sample from the test set.
joshuamcallan
committed
Changed epoch number to 200 to reduce computation time but increases the loss as a result.
joshuamcallan
committed
Added an error rate plot as well as side-byside comparisons of original test images vs reconstructed images. Not certain yet that they will show on github however.
joshuamcallan
committed
Side-by-side comparisons of test images vs reconstructed images as well as loss rate graph.
joshuamcallan
committed
Description of usage of algorithm with this dataset and future datasets.
joshuamcallan
committed
Talked about the train, test, validation split and also the SSIM of each reconstructed image shown. All are above 0.6!
joshuamcallan
committed
All OASIS data added.
joshuamcallan
committed
Fixed path variables for users convenience to alter path for themselves to obtain OASIS data.
joshuamcallan
committed
Deleted a line so the file is class and function based only.
joshuamcallan
committed
Added information about path locations for user to alter themselves as well as the version of python used.
joshuamcallan
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

s46144603 VQVAE OASIS#478

s46144603 VQVAE OASIS#478
joshuamcallan wants to merge 21 commits intoshakes76:topic-recognitionfrom
joshuamcallan:topic-recognition

Commits on Oct 24, 2022

Loaded OASIS dataset from Cdrive and seperated into train, test and validate - each preprocessed by normalising by factoring by 255.

Created a simple encoder for the VQVAE to reduce the amount of paramters in the model.

Added a simple decoder network for the VQVAE.

Commits on Oct 25, 2022

Added VectorQuantization class in order to create a discrete latent space.

Added a vqvae model by feeding inputs into the encoder and the encoder through vq_layer for the quantised vae and finally fed the quantised vae through the decoder for the model.

Added training loop for VQVAE in order to backpropagate and record losses.

Added predicitions using the test set based on the VQVAE model.

Put the training model into the wrong file (predict.py) so swapped it to the train.py file and also added plots of the losses and metrics during training.

Removed the training model from this file and added functionality to compare original vs reconstructed images as well as calculate the structured similarity (SSIM) of these two images).

Added a description of the VQVAE model and how it works. Also added a main title and description of the problem. Some headings were also made.

Added more description for implementation of dataset as well as module dependencies and references.

When choosing the random images, it was selecting from the train dataset so fixed this to sample from the test set.

Changed epoch number to 200 to reduce computation time but increases the loss as a result.

Added an error rate plot as well as side-byside comparisons of original test images vs reconstructed images. Not certain yet that they will show on github however.

Side-by-side comparisons of test images vs reconstructed images as well as loss rate graph.

Description of usage of algorithm with this dataset and future datasets.

Talked about the train, test, validation split and also the SSIM of each reconstructed image shown. All are above 0.6!

All OASIS data added.

Fixed path variables for users convenience to alter path for themselves to obtain OASIS data.

Deleted a line so the file is class and function based only.

Added information about path locations for user to alter themselves as well as the version of python used.

s46144603 VQVAE OASIS#478joshuamcallan wants to merge 21 commits intoshakes76:topic-recognitionshakes76/PatternFlow:topic-recognitionfrom joshuamcallan:topic-recognitionjoshuamcallan/PatternFlow:topic-recognitionCopy head branch name to clipboard

Commits

Commits on Oct 24, 2022

Commits on Oct 25, 2022

s46144603 VQVAE OASIS#478
joshuamcallan wants to merge 21 commits intoshakes76:topic-recognitionfrom
joshuamcallan:topic-recognition