S., Mody, ShrilMody, ShrilS.J., Thakkar, JanviThakkar, JanviJ.2025-09-012025-09-01978145038585597814503143989781450396387978145039001997814503902179781450348270978145038196397814503224859781450348201978145036445410.1145/3493700.34937492-s2.0-85122685482https://www.scopus.com/inward/record.uri?eid=2-s2.0-85122685482&doi=10.1145%2F3493700.3493749&partnerID=40&md5=877dcc3df2675058f201613497d24235https://d8.irins.org/handle/IITG2025/29406In this paper, we analysed the image generation using scene graphs. Three models - SGDET, SG2IM, and Stanford Parser were used for experimental purposes. We evaluated our design pipeline on MS COCO and Visual Genome datasets and compared the images generated using different metrics. � 2022 Elsevier B.V., All rights reserved.EnglishComputer visionImage generationsScene-graphsStanfordThree modelsImage analysisConference paper20220