Research And Implementation Of Agent-based Virtual Interview System

The development of college students' knowledge and qualities and the training of skills improvement are important components of the education to improve the comprehensive abilities of college students.Although college students have a high level of professional knowledge and skills,they still lack many qualities and abilities.Among them,the common problems that directly affect the social life and professional ability of college students include interview anxiety and lack of interview skills.Due to the high cost of training and the shortage of excellent teachers,the existing interview and counseling institutions can not support users to train anytime and anywhere.In response to this situation,combined with virtual reality(VR)and human-computer interaction technology,this paper develops an agent-based virtual interview system for college student interview quality training.In this paper,demand analysis,outline design and detailed design are carried out for the virtual interview system based on agent.In the demand analysis,this paper analyzes the requirements of virtual interview system,introduces the overall situation of virtual interview system,and further describes the functional and non functional requirements of virtual interview system.In the outline design,the main hardware equipment of the virtual interview system based on agent interaction,the architecture of the virtual interview system and the design of the interview operation process are described.In the part of detailed design,the logic design of virtual interview system is introduced firstly,and then the main function design of the system is described.Then this paper realizes the system based on the above analysis and design.Among them,the system uses C#and Unity3D software for system implementation,system interface design and virtual scene design.This article uses SQLite database for the data storage of interview questions and user interview performance.In order to improve the interviewer's ability to answer questions,the psychological quality of the test,subjective experience,and their behavioral performance in the simulation,this article integrates various aspects of appearance,behavior,sound and other characteristics to build a virtual agent with different personality characteristics for interview training.These virtual agents not only have expressive features,but also reactive features,which can provide feedback based on the status of the user,and can also react differently based on the status of the user.In this article,virtual agents are modeled and animated using Maya.The interview system proposed in this paper integrates facial recognition,facial expression recognition,motion recognition,brain-computer interface,active stereoscopic projection and other technologies,and its functions include:system login,multi-mode perception and analysis of physiological signals and behaviors of interviewees,and interview module.Among them,the system login module is completed by using Kinect for face recognition.The main functions of the multi-mode perception and analysis module of physiological signals and behaviors of interviewees include:tracking and recognizing the user's posture with Kinect,and realizing the capture of irregular movements during the interview;Wearable physiological devices are used to collect and analyze the physiological data of users to interpret the user status.Virtual interviewers interact with users according to the user status.The camera is used to capture the video stream of users in real time and detect their facial expressions in real time.The main functions of the interview section include:selection of virtual interviewers with different types of interview scenarios and different personality traits;The interviewer interacts with the user through voice,body and expression.The analysis and feedback of the interview situation of the user,to realize the correction of the non-standard behavior of the user during the interview.Finally,in order to test the effect of the VR interview training system designed on shy individuals' anxiety response and performance improvement,this paper conducted a user test on it,including the evaluation of users' interview anxiety and interview performance.Among them,interview anxiety was evaluated by behavioral and physiological indicators.Interview performance is measured by three measures:average time spent answering each question,adequacy,and fluency.The tests showed that the more the shy college students used the system,the lower their anxiety levels during the interview.In addition,while the average time users spent answering each question did not increase,the adequacy and fluency of their answers increased.Therefore,the virtual reality interview training system can effectively improve the interview skills of Chinese college students and relieve their interview anxiety.