College of Computer Science and Information Technology / University of Kerbala has discussed the M.A. thesis which is entitled “Image Generation Conditioned on Text Using Diffusin Model.”
The study, submitted by Sara Faiz Abdul Ghani, aims at developing a better stable diffusion model to overcome challenges related to this model and produce high-quality images from text.
The study concludes that the proposed method after fine-tuning stable diffusion model leads to a significant improvement in production of diverse images that match the text, and results of improved model indicate a higher Inception Score when compared to basic model.
The study recommends using this model to generate video from text due to the robustness of the diffusion model.