This repository contains the code and pipelines for the project Believing is Seeing: Unobserved Object Detection using Generative Models, which aims to enhance object detection in environments where certain objects are unobserved or occluded. By leveraging generative models, the approach focuses on inferring the presence of objects based on learned priors and contextual information.