Multi-feature-based interview content credibility evaluation method and system

Abstract

The invention discloses a multi-feature-based interview content credibility evaluation method and system. The method comprises the following steps of: collecting interview audio and video data of an interviewer in real time, extracting physiological features, micro-expression features and voice features in real time through combined algorithm processing, fusing the features into feature vectors byusing a multi-modal fusion strategy, inputting the feature vectors into a credibility classification prediction model, acquiring interview content credibility levels in real time for display and reminding, and finally integrating the interview content credibility levels of the interviewer participating in the interview of each time period to obtain the overall credibility evaluation of the interview. The system mainly comprises: a data acquisition unit, a data analysis processing unit, a data model construction unit, an output display unit and a data storage unit. According to the method, a non-contact test evaluation mode is adopted, so that the stress response of the interviewer can be effectively avoided or reduced, a multi-feature fusion strategy is adopted, and the classification prediction model considers the situation of partial feature loss, and the authenticity and reliability of credibility evaluation can be effectively improved.

Description

technical field [0001] The invention relates to the technical field of audio and video information processing, in particular to a method and system for evaluating the credibility of interview content based on multiple features. Background technique [0002] In the process of purposeful interviews such as interviews, negotiations, and interrogations, the interviewees may be motivated to achieve their own goals, such as striving for a higher interview score, obtaining more benefits in the negotiation, shirking responsibility, or evading responsibility. Behaviors that violate the facts such as exaggerating, lying, and covering up. At present, there have been many researches on polygraph detection at home and abroad, and some applications in specific fields such as criminal investigation have been realized, such as multi-channel physiological data and EEG acquisition and analysis technology, micro-expression analysis technology, speech analysis technology, etc. Among them, the ...

AI Technical Summary

Problems solved by technology

The above polygraph techniques all have merits, but they also have their own shortcomings. In addition, because they are basically applied independently, there is no mature fusion strategy and model to further improve the polygraph accuracy.

Images