We designed an Android application which uses image processing, segmentation, HMM and KNN models hosted on a server to translate the gestures and signs of Indian Sign Language used by a hearing- or speech-impaired person.

We, analyse diffusion models and other generative models comparing their generation ability to humans by adapting the 'diversity vs. recognizability' scoring framework from Boutin et al, 2022 as well as comapre human category diagnostic features against those derived from diffusion models.