We are regularly engaged on bettering the precision and remember of your model by experimenting with even more refined networks. We are also working on scaling up outside of the initial trend groups, in addition to increasing to other domains.Soon after several levels of rating the de-duplication stage is executed to eliminate any duplicate visuals