Please use this identifier to cite or link to this item:
http://hdl.handle.net/123456789/18380
Title: | Text to Scene Generator Via Artificial Intelligence |
Authors: | Haider, Wasi |
Keywords: | Computer Sciences |
Issue Date: | 2019 |
Publisher: | Quaid-i-Azam University |
Abstract: | Text to scene generator is a web-based application powered by artificial intelligence. This system will be helpful for story writers and comic book writers to generate scenes for their story. When we tell someone a story or we read it, we picture the scene in our minds but we can’t see it actually. In this proposed system the user will be able to give text input in natural language and then the system will generate a scene according to the objects specified in the sentences given and then using artificial intelligence it will parse the sentence and generate a tree for parent objects to the children objects with their positional relations. For each object in the tree it will place the image of the object to main scene according to the relation found with other objects. Finally the user will be able to save the scene as it will be 2D image. There will be an administrator side which will allow the administrator to add images of the objects and sentences for model training. |
URI: | http://hdl.handle.net/123456789/18380 |
Appears in Collections: | BS |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
COM 2391.pdf | COM 2391 | 1.21 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.