Accessing video games is challenging for people with upper-limb impairments, especially when multiple inputs are required in rapid succession. Human cooperation, where a copilot assists the main player, has been proposed as a solution, but relying on a human assistant poses limitations in terms of availability and co-location. An alternative solution is to use partial automation, where the player is assisted by a software agent. In this work, we present a study with 13 participants with upper-limb impairments, comparatively evaluating how participants collaborate with their copilot in human cooperation and partial automation. The experiment is supported by GamePals, a modular framework that enables both human cooperation and partial automation on existing third-party video games.

翻译：对于上肢障碍人士而言，访问视频游戏具有挑战性，尤其是在需要快速连续执行多输入操作时。人类协作——即由副驾驶协助主要玩家——曾被提出作为解决方案，但依赖人工辅助在可用性和共处性方面存在局限。另一种解决方案是采用部分自动化，即由软件智能体辅助玩家。在本研究中，我们招募了13名上肢障碍参与者，通过对比评估探讨了参与者在人类协作与部分自动化两种模式中如何与协作者配合。实验依托GamePals框架实现，该模块化框架能够在现有第三方视频游戏中同时支持人类协作与部分自动化功能。