Friday 27 October:

• Github, Overleaf, Zotero, HPC, Docs, Drive

Sunday 29 October:

• Read about current existing architectures commonly used for other computer vision tasks: UNet (Nicolo, Daniel), VGGNet(William, Gloria)
• Look for other approaches besides UNet and VGGNet
• Read about the “science background”:
  • De Angelis, S. et al. Three dimensional characterization of nickel coarsening in solid oxide cells via ex-situ ptychographic nano-tomography. Journal of Power Sources 383, 72–79 (2018).(Link)
  • De Angelis, S. et al. Ex-situ tracking solid oxide cell electrode microstructural evolution in a redox cycle by high resolution ptychographic nanotomography. Journal of Power Sources 360, 520–527 (2017). (Link)
• Review theoretical part of the course, week 4 cnn notebook, chapter of book.

Monday 30 October:

We decided on Unet3+.

We went through the documentation.

We did data description

Imported the code into repository (unet code)

We split the data

Friday 03 November:

• Data plotting
• Data Preparation, reshaping pixel size, divided the dataset into training and testing tuples + loaded it into a data loader + transformation from array to tensor