"Teams will train models to progress through a mix of 6 Computer Vision and Audio Processing tasks of increasing difficulty. Tasks will be unlocked as you clear the challenges. You will be scored based on the effectiveness of your models for each challenge, and the best-scoring teams would be given prizes. During the 2-day hackathon, you will have to decide if you would like to push for a higher score for a challenge, or to do well enough to be able to complete all the challenges. May the best teams win!" - DSTA Brainhack TIL 2021
C1, C3, C5 are computer vision tasks, while C2, C4, C6 are audio processing tasks.
This repository contains the archive training code for the DSTA Brainhack TIL Competition 2021. Training was done on a rented GPU on vast.ai. Our team managed to clinch the 1st Runner-Up in the competition.
Not all notebooks were saved back then, but I manage to retrieve one for each task and store them here now (C1_Notebook.ipynb for Computer Vision task and C4_Notebook.ipynb for Audio Processing task).
Slides for the final presentation are also included in the repository (TIL_2021_Slides.pdf).