Training an Agent for FPS Doom Game using Visual Reinforcement Learning and VizDoom
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2017, Vol 8, Issue 12
Abstract
Because of the recent success and advancements in deep mind technologies, it is now used to train agents using deep learning for first-person shooter games that are often outperforming human players by means of only screen raw pixels to create their decisions. A visual Doom AI Competition is organized each year on two different tracks: limited death-match on a known map and a full death-match on an unknown map for evaluating AI agents, because computer games are the best test-beds for testing and evaluating different AI techniques and approaches. The competition is ranked based on the number of frags each agent achieves. In this paper, training a competitive agent for playing Doom’s (FPS Game) basic scenario(s) in a semi-realistic 3D world ‘VizDoom’ using the combination of convolutional Deep learning and Q-learning by considering only the screen raw pixels in order to exhibit agent’s usefulness in Doom is proposed. Experimental results show that the trained agent outperforms average human player and inbuilt game agents in basic scenario(s) where only move left, right and shoot actions are allowed.
Authors and Affiliations
Khan Adil, Feng Jiang, Shaohui Liu, Aleksei Grigorev, B. B. Gupta, Seungmin Rho
Keywords
Related Articles
- EP ID EP251587
- DOI 10.14569/IJACSA.2017.081205
- Views 81
- Downloads 0
A Collaborative Filtering Recommender System Model for Recommending Intervention to Improve Elderly Well-being
In improving elderly well-being nowadays, people at home or health care centre are mostly focusing on guarding and monitoring the elderly using tools, such as CCTV, robots, and other appliances that require a great deal...
Singkat: A Keyword-Based URL Shortener and Click Tracker Package for Django Web Application
In recent years, Python has been gaining popularity as web scripting/programming language. In this research we propose Singkat, an open source uniform resource locator (URL) shortener package with web user interface buil...
Role of Explicit Knowledge Management and Reuse in Higher Educational Environment
Role of knowledge management and knowledge reuse has been investigated analytically in higher educational environment using Nonaka & Takeuchi and Harsh models. It has been observed that in three dimensional environment k...
Smart City Parking Lot Occupancy Solution
In the context of Smart City projects, the management of parking lots is one of the main concerns of local administrations and of industrial solution providers. In this respect, we have presented an image processing appl...
Comparison and Analysis of Different Software Cost Estimation Methods
Software cost estimation is the process of predicting the effort required to develop a software system. The basic input for the software cost estimation is coding size and set of cost drivers, the output is Effort in ter...