Welcome to the Bahria University DSpace digital repository. DSpace is a digital service that collects, preserves, and distributes digital material. Repositories are important tools for preserving an organization's legacy; they facilitate digital preservation and scholarly communication.
dc.contributor.author | Obaid Ullah Khan, 01-134181-098 | |
dc.contributor.author | Muhammad Zahid, 01-134181-086 | |
dc.date.accessioned | 2022-06-16T08:26:42Z | |
dc.date.available | 2022-06-16T08:26:42Z | |
dc.date.issued | 2022 | |
dc.identifier.uri | http://hdl.handle.net/123456789/12837 | |
dc.description | Supervised by Ms. Maryam Bibi | en_US |
dc.description.abstract | GAN is an emerging machine learning technique that has the ability to convert text into a realistic Photo. Realistic-Photo Generator Using Text Description is a combination of NLP and Computer Vision. Combining two data types together to perform this task is a challenging task. However, a recent image captioning project that utilized the multi-model architecture showed that we are capable of completing this task. Image captioning is Image to Text the reverse process of Realistic-Photo Generator Using Text Description. Realistic-Photo Generator Using Text Description can benefit and be useful in the police department to generate a realistic photo of a criminal in a few minutes as in real life sketch of the criminal drawn by an artist consumes a lot of time and effort. The main purpose of this project is to develop web and android applications and train the dataset using GAN to demonstrate suggested applications practically. The application will run over web browsers and android smartphones with internet connectivity. All of the modules that deal with image loading, model saving, photo generation, and user feedback are fully functional. However, we are unable to produce a high-quality photograph, as stated that Stage-1 produces a low-quality image. So, if we want good quality outcomes in generated photos, we must train the Stage-2 model, but our system does have not enough capabilities to train the Stage-2 model, which is why we are unable to generate a high-quality photo | en_US |
dc.language.iso | en | en_US |
dc.publisher | Computer Sciences BUIC | en_US |
dc.relation.ispartofseries | BS (CS);MFN-P 10443 | |
dc.subject | High-Quality Photo. | en_US |
dc.subject | Text Description | en_US |
dc.title | Realistic-Photo Generator Using Text Description. | en_US |
dc.type | Project Reports | en_US |