Track Introduction
Teams participating in this track are required to train their own models to complete challenges within a virtual scenario. The focus is on evaluating foundational decision-making AI development skills, solutions for both single-agent and multi-agent systems, model architecture design, as well as exploring reinforcement learning algorithm design and training methods.
Target Participants
This track is open exclusively to full-time associate degree and undergraduate students enrolled in higher education institutions worldwide. Participants must form a team of 2 to 5 students from the same institution, with no restrictions on their major. Each participant can only be a member of only one team. Furthermore, each team must be guided by 1 to 3 faculty members from the same institution.
Rewards
The prize money for this track is set as above, with amounts in Chinese Yuan (CNY) and pre-tax.
In addition to the prize money, participating teams will also receive the following rewards:
Teams that complete the preliminary round (by successfully submitting a model and passing verification) will receive an official certificate of participation.
Outstanding participants may also be offered internship opportunities related to Tencent’s AI Arena Program, as well as priority access to Tencent Group’s campus recruitment and internship programs. Sepcific details will be notified separately.
Schedule
About the Challenge：In this challenge, participating teams need to use algorithms to train models that drive agents to learn movement strategies through continuous exploration of the map. They must wisely use summoner skills and acceleration boosts to reach the endpoint within a limited time while collecting as many treasure chests as possible.
The map features starting and ending points, roads, obstacles, acceleration boosts, and treasure chests. Agents have a limited local field of view and can move around the map, deploy summoner skills, and acquire rewards contained within the treasure chests. (A comprehensive development guide with more details will be provided on the platform following successful registration.)
Objective：Participating teams must locally train and submit a model within a limited time. Their goal is to control agents on the assessment map to gather as many points as possible in the least amount of time, fulfilling the mission of embarking on a return adventure to the mystic realm.
Ranking Rules： After the submission phase ends, the system will automatically execute the challenge with the latest model submitted by each participating team. Teams will be ranked according to their scores, and these rankings will constitute the final results for this phase of the competition.
Advancement Rules： The top 80 teams on the scoreboard will advance to the next stage.
About the Challenge：IIn this challenge, participating teams are required to train model-driven agents using algorithms. These agents will continuously explore the 1v1 map of the game "Honor of Kings" to learn the optimal strategy, aiming to be the first to destroy the opponent's camp crystal to achieve victory.
The map used in this challenge is elongated, with resurrection points for both teams' agents at each end, and a crystal belonging to each camp positioned in front of the resurrection points. These crystals continuously produce minions for their respective camps, which automatically advance towards the opponent's camp, attacking defense towers, crystals, and heroes along the way. Ahead of the crystals are the camp's defense towers, capable of attacking enemy heroes and minions within range. Agents have the freedom to move around the map and unleash skills at will. (Detailed instructions will be provided in the development guide available on the platform after successful registration)
Objective：Teams are required to utilize the allocated computing resources within a specified time to train their models. The objective is to let these models to learn the optimal winning strategy through continuous exploration of the 1v1 maps, enabling them to secure as many victories as possible in matches against other teams.
Ranking Rules： After the submission phase ends, the system will automatically conduct multiple rounds of matchmaking using the latest model submitted by participating teams. Each team will compete in an equal number of matches against all opponents in their division and will be ranked based on their accumulated points.
In each round of matchmaking between two teams, all heroes from both sides must participate, with each hero facing off against every opposing hero across two rounds. Winning a round earns 1 point, while a loss yields no points.
Advancement Rules： The top 8 teams on the scoreboard will advance to the next stage.
About the Challenge：In this challenge, participating teams need to use algorithms to train models that drive agents to learn movement strategies through continuous exploration of the map. They must wisely use summoner skills and acceleration boosts to reach the endpoint within a set time frame while collecting as many treasure chests as possible.
The map features starting and ending points, roads, obstacles, acceleration boosts, and treasure chests. Agents have a local field of view and can move around the map, deploy summoner skills, and acquire rewards contained within the treasure chests. (A comprehensive development guide with more details will be provided on the platform following successful registration.)
Objective：Participating teams must locally train and submit a model within a given timeframe. Their goal is to control agents on the assessment map to gather as many points as possible in the least amount of time, fulfilling the mission of embarking on a return adventure to the mystic realm.
Ranking Rules： After the submission phase ends, the system will automatically execute the challenge with the latest model submitted by each participating team. Teams will be ranked according to their scores, and these rankings will constitute the final results for this phase of the competition.
Recommendations
It is advised that each team possess at least one computer adhering to the recommended specifications below to establish the local environment for development and training.