This week, we showed off our prototype to pairs of judges. The team had fun showing the work we have accomplished so far, and we also received good feedback. Main piece of feedback we received was to look into transferred learning, which is a machine learning method where a model developed for a task is reused as the starting point for a model on a second task. This could simplify the training process for us since training a model from scratch is very complicated work.

Above is what we showed off for our prototype. We ran stable diffusion XL locally and ran the base models on some preexisting fanatics merch and asked the AI to generate the back view. These were the results which were not entirely accurate. We are hoping that either with a tuned or custom model the results will be much better.