{"title":"Image-dev: An Advance Text to Image AI model","authors":"Manavkumar Patel, Sonal Fatangare, Aryaman Nasare, Abhijeet Pachpute","doi":"10.1109/PuneCon55413.2022.10014718","DOIUrl":null,"url":null,"abstract":"In the recent years, with the rapid growth of Artificial Intelligence, there is increasing interest in Text-to-Image models. High-quality images can be generated with state-of-art text-to-image AI models such as Imagen, DALL.E-2, Draw-Bench. However, these models struggle with generating well aligned images for conflict category and low database. Therefore, Image-dev is a Text-To-Image model that blends TF-IDF(Term Frequency - Inverse Document Frequency) model along with preposition model, to evaluate the relation between the data object. Proposed model output images have an unparalleled level of artistic finish and an added level of language understanding and interpretation further enhance model to produce conflict category images. Image-dev help user's to generate a high-quality, photorealistic images without any pre-context based on GANs, VAEs and diffusion model. Image-dev is based on diffusion model. Diffusion model is more relevant because of its high quality and realistic output generation capacity.","PeriodicalId":258640,"journal":{"name":"2022 IEEE Pune Section International Conference (PuneCon)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE Pune Section International Conference (PuneCon)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PuneCon55413.2022.10014718","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
In the recent years, with the rapid growth of Artificial Intelligence, there is increasing interest in Text-to-Image models. High-quality images can be generated with state-of-art text-to-image AI models such as Imagen, DALL.E-2, Draw-Bench. However, these models struggle with generating well aligned images for conflict category and low database. Therefore, Image-dev is a Text-To-Image model that blends TF-IDF(Term Frequency - Inverse Document Frequency) model along with preposition model, to evaluate the relation between the data object. Proposed model output images have an unparalleled level of artistic finish and an added level of language understanding and interpretation further enhance model to produce conflict category images. Image-dev help user's to generate a high-quality, photorealistic images without any pre-context based on GANs, VAEs and diffusion model. Image-dev is based on diffusion model. Diffusion model is more relevant because of its high quality and realistic output generation capacity.