Abstract: This project is an innovative web-based application that uses advanced AI techniques for image inpainting, a process of filling in selected regions of an image based on a user-provided text prompt. The application utilizes two powerful models: Stable Diffusion and Segment Anything. The Stable Diffusion model is used for the inpainting process, while the Segment Anything model is used to identify and select regions in the image for inpainting. The user interface, built with Gradio, is intuitive and user-friendly. It allows users to upload an image, select regions on the image, and provide a text prompt that guides the inpainting process. The selected regions are then filled in with content that is generated based on the text prompt, creating a unique and personalized result. In addition to the image inpainting feature, the application also includes a vulgarity speech detection mechanism. This feature uses a trained SVM model to analyze the text prompt and detect any offensive or vulgarity speech. If such speech is detected, the application does not proceed with the inpainting process and instead displays a warning message to the user. The application demonstrates the potential of AI in digital art and content creation, providing a tool that is not only functional but also encourages creativity and personal expression. It also underscores the importance of ethical considerations in AI applications, with its inclusion of a vulgarity speech detection feature. Overall, this project represents a significant contribution to the field of AI-powered digital art, offering a unique tool that combines advanced image inpainting techniques with a user-friendly interface and ethical safeguards.
Keywords: Stable Diffusion, Segment Anything Model (SAM), Gradio, Support Vector Machine (SVM), Vulgarity speech, Text prompts, Image inpainting
Cite:
Sahan G Kotian, Shreyas Shettigar, Sharan BS, Vishal M Shettigar, Mrs. Anuburajam.M, "VULGARITY DETECTION IN STABLE DIFFUSION INPAINTING WITH SAM ", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 3, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13338.
|
DOI:
10.17148/IJARCCE.2024.13338
