Double-recording data quality inspection method, device and system

Abstract

The invention discloses a double-recording data quality inspection method, device and system, which can be used in the financial field and other fields. The method comprises the following steps: determining a plurality of segmentation time points corresponding to double-recording data; matching the timestamp information with the segmentation time points, and if the matching is passed, segmenting the video data and the audio data according to the timestamp information; performing confirmation action detection and picture analysis on the segmented video data to obtain a confirmation action detection result and a signature picture group, and performing identification to obtain a signature identification result; performing voice recognition on the segmented audio data to obtain segmented text data, and performing text rule matching to obtain a text matching result; performing validity detection according to a preset validity rule to obtain a validity detection result; and according to the confirmation action detection result, the signature recognition result, the text matching result and the validity detection result, generating a double-record data quality inspection result. According to the invention, the detection accuracy and the double-recording quality inspection detection success rate are improved.

Description

technical field [0001] The invention relates to the technical field of audio and video processing, in particular to a double recording data quality inspection method, device and system. Background technique [0002] In order to protect the rights and interests of consumers, regulatory agencies require commercial banking financial institutions to regulate the sales behavior of financial institutions through audio and video recording (double recording) when selling financial products such as wealth management and consignment insurance policies. At present, commercial banks usually cache video files locally, and upload them to the cloud asynchronously after the entire double-recording video is recorded, for subsequent compliance review by regulatory authorities. [0003] In order to ensure the compliance of double-recorded videos, financial institutions generally use artificial intelligence technology to check offline audio and video data, but there are the following problems, ...

AI Technical Summary

Problems solved by technology

[0003] In order to ensure the compliance of double-recorded videos, financial institutions generally use artificial intelligence technology to check offline audio and video data, but there are the following problems, resulting in a low success rate of quality inspection: (1) The customer confirmation process relies on voice-to-text rules match, but the customer’s voice is low or the environment is noisy, the customer confirmation process is likely to cause the quality inspection to fail

(2) In the link of customer signature action, the action detection model is generally used for recognition, but the recognition accuracy of the action detection model itself is not high. At the same time, when the account manager displays the signature content, sometimes the picture is blurred and the signature content cannot be recognized, resulting in quality inspection. Increased probability of failure

(3) Each link of the speech template relies on voice-to-text to perform segmentation (text classification), and then conducts subsequent corresponding quality inspection content for each segment content according to the template detection points (for example: certificate detection, signature action detection, customer Confirmation, etc.), the accuracy of this segmentation method needs to be improved at present, if there is an error in the segmentation, then subsequent related detection cannot be carried out

Images