VR与AI赋能的沉浸式情境口译教学模式研究

Platform of Immersion Interpreting Teaching in Virtual Reality with AI Assessment:An Exploration towards the Situated Interpreting Learning

为顺应智能环境下的教育发展,充分激发信息技术革命对外语教学的影响,该研究在人工智能(AI)、虚拟现实(VR)和5G通讯新技术赋能下,对口译教学模式进行了全新探索,提出沉浸式情境口译教学模式。该模式针对传统口译教学受时空约束,口译真实应用场景与课堂教学差异较大,人才培养质量与语言服务市场需求尚有差距的痛点,立足于全景视频和3D渲染技术,构建虚拟仿真的口译场景,全方位模拟口译活动中的情境因素,从教学过程、教学方法和教学材料等方面对当前口译教学模式进行革新,帮助提升学生在拟真场景中的口译实践应用能力,并利用AI技术辅助教师进行口译评测,便于开展个性化指导及培养,不仅推动了大数据与外语教育的深度融合,同时为人工智能时代的口译教学及学习研究提供一个全新案例。

Problems persist during the process of second language acquisition, such as lacking face-to-face communication in the target language,and no effective feedback in evaluating and monitoring learners’ performance.Similar problems can also be found in acquiring more complicated second language skills like interpreting.The scholarly agreement was achieved that interpreting is a highly situated activity.Interpreters’ performance is heavily affected by both linguistic factors and extra-4 inguistic factors as well as the environment of the interpreting site.Sufficient knowledge of the themes and subject matters of the interpreting events,and good command of two working languages, may determine the quality of the interpreting performance.Traditionally,interpreter trainers prepare in classroom working sessions guiding students to acquire appropriate strategies and develop relevant skills.However,stress in the marketplace differs from stress in the class-room,not to mention the various kinds of distractions that may happen during the task.In order to overcome the above problems in interpreting training and learning,we attempted to build a comprehensive and interactive training scheme based on the technology of Virtual Reality(VR) and Artificial Intelligence(AI).The paper introduced the design and implementation of such a scheme.i.e., the Platform of Immersion Interpreting Teaching in Virtual Reality with AI Assessment(PIIT).Such attempt was against the backdrop of the ever-deepening pedagogical reform in self-motivated learning and a further integration between education and technology in China,like on line-offline teaching and flipped-classro o m.VR can provide a completely spatialized visualization to language learners,enabling the full presence and interaction of multiple users with input and output from and to the full human sensorium.It also integrates all input and output information in a global information processing system to realize the intercommunication through a full range of intelligent products and environments in real space.It has been widely acknowledged that the integration of VR and AI on distributed platforms can provide a promising framework for language learning technology.We define PIIT as a scenario-based,next-generation virtual platform.Interpreting students can be immersed into VR scenarios simulating interpreting events in real market and interact with the system by high-end VR headsets.PIIT is embedded with three classical virtual scenarios in consecutive interpreting,aiming at improving situated consecutive interpreting skills of student interpreters at both undergraduate and postgraduate levels.This pro posed project also integrates the AI assessment system into the virtual environment.It can evaluate students’ performance based on their reaction time,level of completeness,accuracy,and fluency of the interpreting task.The project supports both visual and auditory channels of interaction,with learner’s interpreted speech as the input,and the output includes graphical scenes and speech from the speaker in the scenario.PIIT was an attempt to explore new ways of interpreting training,by simulating the time and spatial restrictions of a real interpreting site in a physical classroom.It is hoped that PIIT would make a change to teaching procedures,methods as well as materials.Such attempt is to seek further integration of big data and foreign language teaching,enlightening interpreting training and studies in the AI era.