Week 14 (6/12 - 10/12)

Ensuring the OpenFace analysis is run.

Development

Building on what was done last week, my goal was to ensure the OpenFace analysis is run and the subsequent files are uploaded to the correct container in the BLOB storage. This is now fully functional! Building on the flow outlined last week...

1. User uploads an image.

2. The image is uploaded to the BLOB.

3. The Function App is triggered and makes a call containing the container name, media type and analysis type to the OpenFace service.

4. The OpenFace service downloads the corresponding BLOB container.

5. Media inside the container is analysed based on the parameters from step 3, generating a .csv stored in /processed. More information can be found here.

6. The .csv file is uploaded to the BLOB.

7. The Function App is triggered once again. It now sees what kind of file is uploaded. As this is a .csv, it makes a call to the Analysis module.

Now the pipeline is ready for the analysis and reporting modules!

Portfolio

As there is not much time left before the Christmas holidays, it is high time to continue the development of this portfolio. Though there are still a few weeks left, it is important there are multiple meetings planned to receive feedback. The plan as of now is to go over the contents of the portfolio each week leading up to the holidays (two sessions) and one straight after the holidays (the day before the hand-in date). This lets me get some important feedback for during the holidays, and allows me to correct any last-minute mistakes or omissions before it is handed in.

To make sure there is plenty to discuss next week, I continued working on the portfolio in the second half of the week. Working on the portfolio will take priority over development leading up to the break, though I still hope to have the analysis service working by the time the break starts.

Spoilers for next week

I will mainly be working on my portfolio. Development-wise, I will be preparing for the analysis module. This means adding the ability to upload parameter sets from the front-end and preparing a database for the parameters to be stored in.