Domain randomisation enhances the transferability of vision models across visually distinct domains with similar content. However, current methods heavily depend on intricate simulation engines, hampering feasibility and scalability. This paper introduces BehAVE, a video understanding framework that utilises existing commercial video games for domain randomisation without accessing their simulation engines. BehAVE taps into the visual diversity of video games for randomisation and uses textual descriptions of player actions to align videos with similar content. We evaluate BehAVE across 25 first-person shooter (FPS) games using various video and text foundation models, demonstrating its robustness in domain randomisation. BehAVE effectively aligns player behavioural patterns and achieves zero-shot transfer to multiple unseen FPS games when trained on just one game. In a more challenging scenario, BehAVE enhances the zero-shot transferability of foundation models to unseen FPS games, even when trained on a game of a different genre, with improvements of up to 22%. BehAVE is available online at https://github.com/nrasajski/BehAVE.
This paper describes an AI agent that plays the popular first-person-shooter (FPS) video game `Counter-Strike; Global Offensive' (CSGO) from pixel input. The agent, a deep neural network, matches the performance of the medium difficulty built-in AI on the deathmatch game mode, whilst adopting a humanlike play style. Unlike much prior work in games, no API is available for CSGO, so algorithms must train and run in real-time. This limits the quantity of on-policy data that can be generated, precluding many reinforcement learning algorithms. Our solution uses behavioural cloning - training on a large noisy dataset scraped from human play on online servers (4 million frames, comparable in size to ImageNet), and a smaller dataset of high-quality expert demonstrations. This scale is an order of magnitude larger than prior work on imitation learning in FPS games.
The emerging progress of eSports lacks the tools for ensuring high-quality analytics and training in Pro and amateur eSports teams. We report on an Artificial Intelligence (AI) enabled solution for predicting the eSports player in-game performance using exclusively the data from sensors. For this reason, we collected the physiological, environmental, and the game chair data from Pro and amateur players. The player performance is assessed from the game logs in a multiplayer game for each moment of time using a recurrent neural network. We have investigated that attention mechanism improves the generalization of the network and provides the straightforward feature importance as well. The best model achieves ROC AUC score 0.73. The prediction of the performance of particular player is realized although his data are not utilized in the training set. The proposed solution has a number of promising applications for Pro eSports teams and amateur players, such as a learning tool or a performance monitoring system.
Increasing the scale of reinforcement learning experiments has allowed researchers to achieve unprecedented results in both training sophisticated agents for video games, and in sim-to-real transfer for robotics. Typically such experiments rely on large distributed systems and require expensive hardware setups, limiting wider access to this exciting area of research. In this work we aim to solve this problem by optimizing the efficiency and resource utilization of reinforcement learning algorithms instead of relying on distributed computation. We present the "Sample Factory", a high-throughput training system optimized for a single-machine setting. Our architecture combines a highly efficient, asynchronous, GPU-based sampler with off-policy correction techniques, allowing us to achieve throughput higher than $10^5$ environment frames/second on non-trivial control problems in 3D without sacrificing sample efficiency. We extend Sample Factory to support self-play and population-based training and apply these techniques to train highly capable agents for a multiplayer first-person shooter game. The source code is available at https://github.com/alex-petrenko/sample-factory
eSports is the rapidly developing multidisciplinary domain. However, research and experimentation in eSports are in the infancy. In this work, we propose a smart chair platform - an unobtrusive approach to the collection of data on the eSports athletes and data further processing with machine learning methods. The use case scenario involves three groups of players: `cyber athletes' (Monolith team), semi-professional players and newbies all playing CS:GO discipline. In particular, we collect data from the accelerometer and gyroscope integrated in the chair and apply machine learning algorithms for the data analysis. Our results demonstrate that the professional athletes can be identified by their behaviour on the chair while playing the game.
Today's competition between the professional eSports teams is so strong that in-depth analysis of players' performance literally crucial for creating a powerful team. There are two main approaches to such an estimation: obtaining features and metrics directly from the in-game data or collecting detailed information about the player including data on his/her physical training. While the correlation between the player's skill and in-game data has already been covered in many papers, there are very few works related to analysis of eSports athlete's skill through his/her physical behavior. We propose the smart chair platform which is to collect data on the person's behavior on the chair using an integrated accelerometer, a gyroscope and a magnetometer. We extract the important game events to define the players' physical reactions to them. The obtained data are used for training machine learning models in order to distinguish between the low-skilled and high-skilled players. We extract and figure out the key features during the game and discuss the results.
Effective and intelligent exploration has been an unresolved problem for reinforcement learning. Most contemporary reinforcement learning relies on simple heuristic strategies such as $\epsilon$-greedy exploration or adding Gaussian noise to actions. These heuristics, however, are unable to intelligently distinguish the well explored and the unexplored regions of state space, which can lead to inefficient use of training time. We introduce entropy-based exploration (EBE) that enables an agent to explore efficiently the unexplored regions of state space. EBE quantifies the agent's learning in a state using merely state-dependent action values and adaptively explores the state space, i.e. more exploration for the unexplored region of the state space. We perform experiments on a diverse set of environments and demonstrate that EBE enables efficient exploration that ultimately results in faster learning without having to tune any hyperparameter. The code to reproduce the experiments is given at \url{https://github.com/Usama1002/EBE-Exploration} and the supplementary video is given at \url{https://youtu.be/nJggIjjzKic}.
Learning robust value functions given raw observations and rewards is now possible with model-free and model-based deep reinforcement learning algorithms. There is a third alternative, called Successor Representations (SR), which decomposes the value function into two components -- a reward predictor and a successor map. The successor map represents the expected future state occupancy from any given state and the reward predictor maps states to scalar rewards. The value function of a state can be computed as the inner product between the successor map and the reward weights. In this paper, we present DSR, which generalizes SR within an end-to-end deep reinforcement learning framework. DSR has several appealing properties including: increased sensitivity to distal reward changes due to factorization of reward and world dynamics, and the ability to extract bottleneck states (subgoals) given successor maps trained under a random policy. We show the efficacy of our approach on two diverse environments given raw pixel observations -- simple grid-world domains (MazeBase) and the Doom game engine.
The recent advances in deep neural networks have led to effective vision-based reinforcement learning methods that have been employed to obtain human-level controllers in Atari 2600 games from pixel data. Atari 2600 games, however, do not resemble real-world tasks since they involve non-realistic 2D environments and the third-person perspective. Here, we propose a novel test-bed platform for reinforcement learning research from raw visual information which employs the first-person perspective in a semi-realistic 3D world. The software, called ViZDoom, is based on the classical first-person shooter video game, Doom. It allows developing bots that play the game using the screen buffer. ViZDoom is lightweight, fast, and highly customizable via a convenient mechanism of user scenarios. In the experimental part, we test the environment by trying to learn bots for two scenarios: a basic move-and-shoot task and a more complex maze-navigation problem. Using convolutional deep neural networks with Q-learning and experience replay, for both scenarios, we were able to train competent bots, which exhibit human-like behaviors. The results confirm the utility of ViZDoom as an AI research platform and imply that visual reinforcement learning in 3D realistic first-person perspective environments is feasible.
Advances in deep reinforcement learning have allowed autonomous agents to perform well on Atari games, often outperforming humans, using only raw pixels to make their decisions. However, most of these games take place in 2D environments that are fully observable to the agent. In this paper, we present the first architecture to tackle 3D environments in first-person shooter games, that involve partially observable states. Typically, deep reinforcement learning methods only utilize visual input for training. We present a method to augment these models to exploit game feature information such as the presence of enemies or items, during the training phase. Our model is trained to simultaneously learn these features along with minimizing a Q-learning objective, which is shown to dramatically improve the training speed and performance of our agent. Our architecture is also modularized to allow different models to be independently trained for different phases of the game. We show that the proposed architecture substantially outperforms built-in AI agents of the game as well as humans in deathmatch scenarios.