Label Buddy 2.0: Automated audio-tagging using transfer learning

This project is an enhancement to the previous work that has been done. Its goal is to make annotation simple and easy while also providing a well-defined manager-nnotator-reviewer framework. The goal of this project is to use Transfer Learning (TL) approaches to make the annotation process easier for the user by offering label predictions. This way, it will be possible to accomplish more with less data and effort. - An analysis of pros and cons of existing work such as BAT, YAMNet and MusicNN. This task will help us understand the drawbacks of the existing implementations and try to avoid them while keeping all the already made advantages. - Enhance the already made database which will be used to store the corrections in order to refine the model. - Research and design of the transfer learning model and an explanation for the architecture. - Implementation of the aforementioned model. - Enhancements of the already made Graphical User Interface providing Mock-up screens. This task includes the design of every addition on every screen in our tool and it should be discussed with the mentors before the implementation. - Implementation of the changes on User Interface and system for the use of the AI-Assistance. - Implementation of the whole AI-Assisting model. - Perform Django lazy loading, testing and Dockerization - Documentation