Mohammad Moaz Tahir

Academic

Career

Verbatim text

GameGini – Game scene generation on input or generated story
GameGini uses Natural Language Processing and a custom-made mapping and parsing mechanism that maps entities and objects from a story, and generates a game world with simple player interactions and object placements synonymous to the story. The story is either based on an input, or it may also be auto-generated using an input prompt.
Features include:
Proper positional placement; Placement of objects synonymous to story description.
Multiple asset chaining; Placement of objects based on the relative positioning amongst themselves in the story.
Simple mechanics; Game mechanics such as day/night time and simple event triggers.

Muhammad Saif Sadiq i191877
Mohammad Moaz Tahir i191904
Mohammad Saud Tahir i191905
Supervisor: Dr. Mirza Omer Beg

Objective
An interactive experience, which on each playthrough, generates a small, unique story, and based on that, generates a unique world to experience it.

Pipeline
Story generation → Story validation → Story parsing → Mapping → Game logic implementation → Game level generation → User plays the game

Timeline
Iteration I (Sep-Oct): Text to object mapping, Entity assets specification, Primitive world generation
Iteration II (Nov-Dec): Story generation, Entity mapping specification, Story parsing logic creation
Iteration III (Jan-Mar): Robust world generation, Game mechanics implementation
Iteration IV (Apr-May): Story parsing and entity mapping enhancements, Objects interaction implementation

Tools and Technologies
Unreal Engine, Python, NLP, Quixel Bridge

Technology Used:
Unreal Engine, Python, NLTK, Quixel Bridge
Supervisor Name:
Dr. Mirza Omer Beg
Group Members:
Muhammad Saif Sadiq (i19 - 1877)
Mohammad Moaz Tahir (i19 - 1904)
Mohammad Saud Tahir (i19 - 1905)

AI enrichment