Sprint Summary

Starting off the project with the first sprint, the team's focus was to get aquainted with Neural Netoworks in Python, as well as Android app development. This sprint's simple goal was to have a preliminary model trained and a first prototype of the app by the end of the week. After having chosen the adequate frameworks, namely Tensorflow and Android Studio, we managed to achieve our objective by the end of the week, even though the solutions for both front and back end were quite rudimentary and especially the AI-model needed to be improved. 

The goal for the second sprint was to get a full working version of the app, which required loading the model onto a server and computing the inference, as well as setting up the connection from the app client to the server using a Java Socket and ultimately, parsing the .json file. This was, unfortunately, not accomplished because of problems and errors on the client side but still, good progress was noticed. To illustrate this, we had a fully working version of the app, while in comparison to the previous week, the connection and the model also became better.

As the second sprint's goal was not met, it was carried over into the third one. In this slightly shorter sprint we got a fully working version to run and tested it with a still preliminary model, which in turn led to reaching our objective for the week. Apart from that, we extracted and collected samples, as we had to balance out the database, which represents our basis for the AI accent detection. We also decided to build from scratch an iOS version of the app in addition to the Android version in order to be available on more devices. 

Last but not least, the goal for the final spint consisted in bringing more features to the app framework by integrating a share button for the obtained results, improving the record microphone button and finalizing the iOS app. Naturally, more samples were continuously gathered to so that the AI model was improved even more and in the end the status of the week's Sprint objective was partially completed as the iOS version of the app is still in progress.