Games are considered a suitable and standard benchmark for checking the performance of artificial intelligence-based algorithms in terms of training, evaluating, and comparing the performance of AI agents. In this research, an application of the Intrinsic Curiosity Module (ICM) and the Asynchronous Advantage Actor–Critic (A3C) algorithm is explored using action games. Having been proven successful in several gaming environments, its effectiveness in action games is rarely explored. Providing efficient learning and adaptation facilities, this research aims to assess whether integrating ICM with A3C promotes curiosity-driven explorations and adaptive learning in action games. Using the MAME Toolkit library, we interface with the game environments, preprocess game screens to focus on relevant visual elements, and create diverse game episodes for training. The A3C policy is optimized using the Proximal Policy Optimization (PPO) algorithm with tuned hyperparameters. Comparisons are made with baseline methods, including vanilla A3C, ICM with pixel-based predictions, and state-of-the-art exploration techniques. Additionally, we evaluate the agent’s generalization capability in separate environments. The results demonstrate that ICM and A3C effectively promote curiosity-driven exploration in action games, with the agent learning exploration behaviors without relying solely on external rewards. Notably, we also observed an improved efficiency and learning speed compared to baseline approaches. This research contributes to curiosity-driven exploration in reinforcement learning-based virtual environments and provides insights into the exploration of complex action games. Successfully applying ICM and A3C in action games presents exciting opportunities for adaptive learning and efficient exploration in challenging real-world environments.

Virtual reality driving simulators have been increasingly used for training purposes, but they are still lacking effective driver assistance features, and poor use of user interface (UI) and guidance systems leads to users’ performance being affected. In this paper, we investigate image–arrow aids in a virtual reality driving simulator (VRDS) that enables trainees (new drivers) to interpret instructions according to the correct course of action while performing their driving task. Image–arrow aids consist of arrows, texts, and images that are separately rendered during driving in the VRDS. A total of 45 participants were divided into three groups: G1 (image–arrow aids), G2 (audio and textual aids), and G3 (arrows and textual aids). The results showed that G1 (image–arrow guidance) achieved the best performance, with a mean error rate of 8.1 (SD = 1.23) and a mean completion time of 3.26 min (SD = 0.56). In comparison, G2 (audio and textual aids) had a mean error rate of 10.8 (SD = 1.31) and completion time of 4.49 min (SD = 0.67), while G3 (arrows and textual aids) had the highest error rate (18.4, SD = 1.43) and longest completion time (6.51 min, SD = 0.68). An evaluation revealed that the performance of G1 is significantly better than that of G2 and G3 in terms of performance measures (errors + time) and subjective analysis such as usability, easiness, understanding, and assistance.