This is the project repo for our team in 10707 Advanced Deep Learning at Carnegie Mellon University in Spring 2022. This project studies the nocaps dataset.
The language modeling for caption evaluation method is in folder 10707-nocaps-master, contributed by @ReedyHarbour.
The scene graph generation (SGG) method is in folder Updown_scenegraph, contributed by @JiahengHu.
The separate scene and object representation (SSO) method is in folder updown-sso-master, contributed by @yuchen-xu.