CIOTechie_ArtificialIntelligence_The-Obstacle-Tower-Challenge_Unity

Unity Introduces The Obstacle Tower Challenge To Test Capabilities Of Artificial Intelligence Software

Artificial Intelligence News

Unity_Introduces-The-Obstacle-Tower-Challenge-To-Test-Capabilities-Of-Artificial-Intelligence-Software Unity Introduces The Obstacle Tower Challenge To Test Capabilities Of Artificial Intelligence SoftwareVideo game engine, Unity has launched a video game challenge, known as the Obstacle tower challenge that will test the vision, control, planning, and generalization capabilities of Artificial Intelligence software. The Obstacle Tower Challenge will utilize a game-like ecosystem of platform-style gameplay with puzzles and planning issues inside a tower setting maneuvering almost 100 floors. The challenge will look at how AI software works in computer vision, locomotion skills, and high-level planning.

As the challenge will begin from February 11 and run through May 24, participants can review all the rules and regulations, download the Starter Kit and start training their agents. In the first round, participants can play up to 25 floors of the Obstacle Tower that will run through March 31. The winners of the first round will proceed to round 2 which will have 100 floors, after that the winners will be announced on June 14. Through the Obstacle Tower Challenge, participants will have the opportunity to become the winner of the prizes, includes cash, travel vouchers, and Google Cloud Platform credits which worth up to USD 100,000. Vice President of AI and Machine Learning at Unity Technologies, Danny Lange stated that each of the tower floors is procedurally-generated that means an AI agent must not only be able to solve a single version of the Tower but any arbitrary version as well. In this way, the company is testing the generalization ability of agents, a key capability that has not usually been examined by benchmarks in the past.

The main purpose of the Obstacle Tower Challenge is to bring up new AI research and solve new problems in reinforcement learning. Artificial Intelligence has been making great progress in conquering high-profile games. Recently, Google DeepMind’s AI AlphaStar defeated StarCraft II pros TLO and MaNa and won 10-1 against the gamers.