Automatic detection and recording system and method based on fusion host terminal AI computing power

Abstract

The present invention discloses an automatic detection video recording system and method based on the AI ​​computing power of the fusion host terminal, and relates to the technical field of automatic video recording detection. The present invention includes a communication analysis module for accepted cases, a mediation strategy generation module, an automatic video recording detection module, and a mediation strategy determination module; the communication analysis module for accepted cases is used to predict the mediation index of each mediation conflict point in the accepted cases; the mediation strategy generation module is used to determine the target adjustment strategy; the automatic video recording detection module is used to automatically detect the video clipping time period; the mediation strategy determination module is used to determine the mediation strategy for the accepted cases. The present invention analyzes the real-time multi-dimensional images in the trial process by integrating the host terminal, and realizes the automatic detection of each clipping time period of the video uploaded to the cloud by the live broadcast server, ensuring that after the trial, the mediation process of each mediation conflict point can be automatically detected and identified synchronously.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic video recording detection, and in particular to an automatic video recording detection system and method based on the fusion of host terminal AI computing power. Background Art

[0002] Civil and commercial cases refer to cases or disputes involving social relations regulated by civil and commercial laws. The four main ways to resolve civil and commercial cases are reconciliation, mediation, arbitration and civil litigation.

[0003] In the existing technology, it is impossible to implement artificial intelligence technology in the mediation practice of civil and commercial cases, that is, to mediate through mediators. During the mediation process, the mediators need to understand the details of the case and formulate case mediation strategies based on the conversations between the parties. This process takes a lot of time and increases the workload of the mediators. In the litigation stage of civil and commercial cases, relevant personnel are unable to automatically crop and mark the video recordings of the trial in real time, resulting in the need for relevant personnel to call up the video recordings afterwards to achieve the above purpose, which is not conducive to the rapid closure of the case. Summary of the invention

[0004] The purpose of the present invention is to provide an automatic detection and recording system and method based on the fusion of host terminal AI computing power to solve the problems raised in the prior art.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an automatic detection and recording system based on the fusion host terminal AI computing power, the system includes a case acceptance communication analysis module, a mediation strategy generation module, a video automatic detection module and a mediation strategy determination module;

[0006] The communication and analysis module for accepted cases is used to obtain the case content uploaded by the applicant of the accepted case to the case processing terminal. The intelligent terminal analyzes the litigation request and evidence content in the acquired case content and generates mediation ideas. The mediator of the accepted case communicates with the applicant and the respondent of the accepted case according to the generated mediation ideas. Based on the communication results, the target integration content is obtained, and the mediation index of each mediation conflict point in the accepted case is predicted;

[0007] The mediation strategy generation module preliminarily generates a mediation strategy based on the mediation index of each mediation conflict point in the pending case, and determines the target adjustment strategy in combination with the professional knowledge and experience of the mediator of the pending case;

[0008] When the applicant of the pending case and the respondent of the pending case do not accept the target mediation strategy, the automatic video detection module analyzes the real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case in the trial through the fusion host terminal, and predicts the shutdown time of the digital trial live broadcast server through the fusion host terminal according to the analysis results, and automatically detects the video clipping time period;

[0009] The mediation strategy determination module is used to determine the mediation strategy for the case to be accepted based on the automatic video recording content of the digital court trial editing and live broadcast server.

[0010] Furthermore, the accepted case communication analysis module includes a mediation idea generation unit, an integration screening unit and a mediation index prediction unit;

[0011] The mediation idea generation unit obtains the case content uploaded by the applicant of the case to be accepted to the case processing terminal, and the intelligent terminal judges whether the litigation request or evidence content in the case content complies with the regulations, and generates a mediation idea based on the judgment result;

[0012] The integration screening unit receives the mediation ideas generated by the mediation idea generation unit, collects the communication content between the mediator of the case to be accepted and the applicant and the respondent of the case to be accepted based on the received mediation ideas, and screens the case content in combination with the litigation request and evidence content in the case content. The specific screening and processing method is: if the supplementary content B of the applicant and the respondent of the case to the evidence content A that does not comply with the regulations still does not comply with the regulations, then the evidence content A is removed from the case content of the case to be accepted; if the supplementary content B of the applicant and the respondent of the case to the evidence content A that does not comply with the regulations complies with the regulations, then the evidence content A is removed from the case content of the case to be accepted, and the supplementary content B is added to the case content of the case to be accepted to obtain the target integrated content;

[0013] The mediation index prediction unit receives the target integration content transmitted by the integration screening unit. The intelligent terminal determines the set of mediation conflict points of the case to be accepted based on the received target integration content, and predicts the mediation index of each mediation conflict point based on the communication results between the mediator of the case to be accepted and the applicant and respondent of the case to be accepted.

[0014] Furthermore, the specific method for the mediation idea generation unit to generate the mediation idea is: if the litigation request and the evidence content in the case content are in compliance with the regulations, then the generated mediation idea is to make the respondent of the case to be accepted recognize the litigation request of the applicant of the case to be accepted; if the evidence content in the case content does not comply with the regulations, then the generated mediation idea is to guide the applicant of the case to be accepted and the respondent of the case to be accepted to supplement the evidence content that does not comply with the regulations; if the litigation request in the case content does not comply with the regulations, then the generated mediation idea is to guide the applicant of the case to be accepted to recognize the regulations corresponding to the litigation request.

[0015] Furthermore, the specific method for the mediation index prediction unit to predict the mediation index of each mediation conflict point is:

[0016] The judicial interpretations related to each mediation conflict point are obtained, and the viewpoints supported by each related judicial interpretation are clarified. Combined with the communication between the mediator of the pending case and the applicant and the respondent of the pending case on each mediation conflict point, the mediation index of each mediation conflict point is predicted. The specific prediction formula is:

[0017] ;

[0018] Among them, i=1,2,…,m, represents the number corresponding to each mediation conflict point, m represents the total number of mediation conflict points in the pending cases, and u i represents the number of judicial interpretations associated with the mediation conflict point numbered i that are the same as the view supported by the applicant of the pending case, U i represents the total number of judicial interpretations associated with the mediation conflict point numbered i, e is a constant and e>1, T 1i 、T 2i represents the communication time between the mediator of the pending case and the applicant of the pending case, and between the mediator of the pending case and the respondent of the pending case on the mediation conflict point numbered i, t 1i ,t 2i W represents the time that the applicant and the respondent of the pending case support their own opinions in the process of communicating with the mediator of the pending case on the mediation conflict point numbered i, respectively. i Represents the mediation index of the mediation conflict point numbered i.

[0019] Furthermore, the mediation strategy generation module obtains the mediation index of each mediation conflict point in the pending case predicted by the mediation index prediction unit. i ≥0.7, it means that the mediation strategy for mediation conflict point numbered i is: support the applicant's view on mediation conflict point numbered i. If 0≤W i<0.7, it means that the mediation strategy for the mediation conflict point numbered i is: use the compromise method to mediate the mediation conflict point numbered i, and the mediator of the case to be accepted chooses whether to adjust the mediation strategy initially generated by the intelligent terminal based on professional knowledge and experience, and obtains the target mediation strategy based on the adjustment result.

[0020] Further, the video automatic detection module includes a multi-dimensional image analysis unit, a live broadcast server shutdown time prediction unit and a video automatic detection unit;

[0021] When any one of the applicant of the pending case and the respondent of the pending case does not accept the target mediation strategy, the multi-dimensional image analysis unit mediates the pending case through a court trial. During the court trial, the fusion host terminal collects real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case, and predicts the initial mediation time and the mediation end time of each mediation conflict point in the pending case based on the collected real-time multi-dimensional images;

[0022] The live broadcast server shutdown time prediction unit predicts the shutdown time of the live broadcast server according to the mediation end time of each mediation conflict point;

[0023] The automatic video detection unit automatically detects each cropping time period of the video uploaded to the cloud by the live broadcast server according to the prediction result of the multi-dimensional image analysis unit, and automatically crops the video uploaded to the cloud by the live broadcast server according to the detection result.

[0024] Furthermore, the specific method for the multi-dimensional image analysis unit to predict the initial mediation time and mediation end time of each mediation conflict point in the accepted case is:

[0025] Based on the real-time multi-dimensional images collected by the fusion host terminal, the real-time voice and text contents of the judge, the applicant of the case to be accepted, and the applicant of the case to be accepted are collected. According to the collected real-time voice and text contents, a real-time voice and text content set of each case-related person is generated, and the initial mediation time point and the mediation end time point of each mediation conflict point in the case to be accepted are analyzed. The specific analysis method is: in each generated real-time voice and text content set, the time when the text content of the mediation conflict point numbered i appears in each voice and text content set, and the total number of times the text content of the mediation conflict point numbered i appears are collected, and according to For S ij Predict the probability value of the initial mediation time for the mediation conflict point numbered i;

[0026] Where c=1,2,…,C, represents the numbering of the number of times each mediation conflict point appears in the generated real-time voice text content set in chronological order, C represents the total number of times each mediation conflict point appears in the generated real-time voice text content set, j=1,2,…,C and j≠c, S ic represents the time when the text content of the mediation conflict point numbered i appears for the cth time in the generated real-time voice text content set, S ij represents the time when the text content of the mediation conflict point numbered i appears for the jth time in the generated real-time voice text content set, F 1ijc 、F 2ijc 、F 3ijc Respectively expressed in [S ij ,S ic ] The matching degree between the real-time voice text content set of the judge, the applicant of the pending case, and the respondent of the pending case generated within the time period and the mediation conflict point numbered i;

[0027] like , then it means S ij is the initial mediation time of the mediation conflict point numbered i, if , then it means S ij It is not the initial mediation time of the mediation conflict point numbered i, X represents a constant and 0.5≤X≤0.7;

[0028] when When S ij The mediation end time of the mediation conflict point numbered i.

[0029] Furthermore, the specific method for the live broadcast service shutdown time prediction unit to predict the shutdown time of the live broadcast server is:

[0030] In the time period [d, f], the mediation end time of the mediation conflict point numbered i is obtained. If the total number of mediation end times obtained is less than m, it means that time f is not the shutdown time of the live broadcast server. If the total number of mediation end times obtained is equal to m, it means that time f+g is the shutdown time of the live broadcast server, where d represents the startup time of the live broadcast server, and g represents the average closing time of the trial obtained based on historical data;

[0031] The specific method of the automatic video detection unit automatically detecting each cropping time period of the video uploaded to the cloud by the live broadcast server is:

[0032] The initial mediation time and mediation end time of each mediation conflict point are obtained to obtain the mediation time period of each mediation conflict point. Within the [d, f] time period, the video corresponding to each mediation time period in the video uploaded to the cloud by the live broadcast server is automatically cropped.

[0033] Furthermore, the mediation strategy determination module obtains the mediation end time of each mediation conflict point, determines the maximum value of the obtained mediation end time, and records the determined maximum value of the mediation end time as R. The video uploaded to the cloud by the live broadcast server within the time period corresponding to the [R, f+g] time period is automatically cropped, and the voice text content in the cropped video is used as the final mediation strategy for the case to be accepted.

[0034] The automatic detection and recording method based on the integration of host terminal AI computing power includes:

[0035] S10: The case content uploaded by the applicant of the pending case to the case processing terminal is obtained, and the intelligent terminal analyzes the litigation request and evidence content in the obtained case content to generate a mediation idea. The mediator of the pending case communicates with the applicant and the respondent of the pending case based on the generated mediation idea, obtains the target integration content based on the communication results, and predicts the mediation index of each mediation conflict point in the pending case;

[0036] S20: Based on the mediation index of each mediation conflict point in the pending cases, a preliminary mediation strategy is generated, and the target adjustment strategy is determined in combination with the professional knowledge and experience of the mediators of the pending cases;

[0037] S30: When the applicant of the pending case and the respondent of the pending case do not accept the target mediation strategy, the real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case in the trial are analyzed through the fusion host terminal, and according to the analysis results, the shutdown time of the digital trial live broadcast server is predicted through the fusion host terminal, and the video clipping time period is automatically detected;

[0038] S40: Determine the mediation strategy for the case to be accepted based on the automatic video content of the digital court trial live broadcast server.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. The present invention provides mediation ideas for mediators of pending cases through intelligent terminals. Based on the communication content between the mediators of pending cases and the parties to the case on adjustment ideas, the mediation index of each mediation conflict point in the pending cases is predicted, and then the target mediation strategy for the pending cases is generated. This process reduces the time for mediators of pending cases to understand and sort out the cases, and further improves the work efficiency of mediators of pending cases.

[0041] 2. The present invention analyzes the real-time multi-dimensional images during the trial by integrating the host terminal, and realizes automatic detection of each cropped time period of the video uploaded to the cloud by the live broadcast server, so as to ensure that after the trial, the mediation process of each mediation conflict point can be automatically detected and identified synchronously, and the judge can retrieve the mediation process of each mediation conflict point at any time when making the closing statement, which further improves the system's automatic detection capability for the video and improves the case closing rate of pending cases.

[0042] 3. When the present invention automatically detects the video, it recognizes the real-time voice and text content in the real-time multi-dimensional image. Based on the recognition result, the matching degree between each real-time voice and text content set and each mediation conflict point, as well as the appearance time of the text content of each mediation conflict point, realizes real-time calculation of the initial mediation time and the mediation end time of each mediation conflict point, thereby improving the intelligent effect of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the working principle structure of the automatic detection and recording system and method based on the integration of the host terminal AI computing power of the present invention;

[0044] Figure 2 The figure is a schematic diagram of the workflow of the automatic detection and recording system and method based on the integration of the AI ​​computing power of the host terminal of the present invention. DETAILED DESCRIPTION

[0045] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0046] Example: Figure 1 and Figure 2 As shown, the present invention provides a technical solution of an automatic detection and recording system and method based on the fusion host terminal AI computing power, an automatic detection and recording system based on the fusion host terminal AI computing power, the system includes a case acceptance communication analysis module, a mediation strategy generation module, a video automatic detection module and a mediation strategy determination module;

[0047] The communication and analysis module for accepted cases is used to obtain the case content uploaded by the applicant of the accepted case to the case processing terminal. The intelligent terminal analyzes the litigation request and evidence content in the acquired case content. The analysis refers to the use of legal logic and reasoning based on specific legal provisions and combined with the facts of the case to draw conclusions on issues such as the application of law, the relationship between rights and obligations, and generate mediation ideas. The mediator of the accepted case communicates with the applicant and the respondent of the accepted case based on the generated mediation ideas. Based on the communication results, the target integration content is obtained, and the mediation index of each mediation conflict point in the accepted case is predicted;

[0048] The case communication analysis module includes a mediation idea generation unit, an integration screening unit, and a mediation index prediction unit;

[0049] The mediation idea generation unit obtains the case content uploaded by the applicant of the case to be accepted to the case processing terminal, and the intelligent terminal judges whether the litigation request or evidence content in the case content complies with the regulations. The regulations refer to the contents stipulated in the normative documents formulated and adopted by the legislative body and have legal effect, and generates mediation ideas based on the judgment results. The specific method is: if the litigation request and evidence content in the case content comply with the regulations, the generated mediation idea is to make the respondent of the case to be accepted recognize the litigation request of the applicant of the case to be accepted; if the evidence content in the case content does not comply with the regulations, the generated mediation idea is to guide the applicant of the case to be accepted and the respondent of the case to be accepted to supplement the evidence content that does not comply with the regulations; if the litigation request in the case content does not comply with the regulations, the generated mediation idea is to guide the applicant of the case to be accepted to recognize the regulations corresponding to the litigation request;

[0050] The integration and screening unit receives the mediation ideas generated by the mediation idea generation unit, collects the communication content between the mediator of the case to be accepted and the applicant and the respondent of the case to be accepted based on the received mediation ideas, and screens the case content in combination with the litigation request and evidence content in the case content. The specific screening and processing method is: if the supplementary content B of the applicant and the respondent of the case to the evidence content A that does not comply with the regulations still does not comply with the regulations, then the evidence content A is removed from the case content of the case to be accepted; if the supplementary content B of the applicant and the respondent of the case to the evidence content A that does not comply with the regulations complies with the regulations, then the evidence content A is removed from the case content of the case to be accepted, and the supplementary content B is added to the case content of the case to be accepted to obtain the target integrated content, which refers to the litigation request in the case content of the case to be accepted, as well as the retained and added evidence content;

[0051] The mediation index prediction unit receives the target integration content transmitted by the integration and screening unit. The intelligent terminal determines the set of mediation contradiction points of the case to be accepted based on the received target integration content. The mediation contradiction point refers to the different opinions of the two parties to the case in the litigation request or evidence content. For example, in a private loan dispute, the transfer record provided by the plaintiff did not specify in detail whether the defendant's repayment belonged to the principal or interest. The plaintiff believed that the defendant's repayment was interest, and the defendant believed that the repayment was principal. Then, whether the defendant's repayment belonged to the principal or interest was a mediation contradiction point. Combined with the communication results between the mediator of the case to be accepted and the applicant and the respondent of the case to be accepted, the mediation index of each mediation contradiction point is predicted. The specific method is:

[0052] The judicial interpretations related to each mediation conflict point are obtained, and the viewpoints supported by each related judicial interpretation are clarified. The related judicial interpretation refers to the content corresponding to the legal obligations and rights related to the mediation conflict point. Combined with the communication between the mediator of the pending case and the applicant and the respondent of the pending case on each mediation conflict point, the mediation index of each mediation conflict point is predicted. The specific prediction formula is:

[0053] ;

[0054] Among them, i=1,2,…,m, represents the number corresponding to each mediation conflict point, m represents the total number of mediation conflict points in the pending cases, and u i represents the number of judicial interpretations associated with the mediation conflict point numbered i that are the same as the view supported by the applicant of the pending case, U i represents the total number of judicial interpretations associated with the mediation conflict point numbered i, e is a constant and e>1, T 1i 、T 2i represents the communication time between the mediator of the pending case and the applicant of the pending case, and between the mediator of the pending case and the respondent of the pending case on the mediation conflict point numbered i, t 1i ,t 2i W represents the time that the applicant and the respondent of the pending case support their own opinions in the process of communicating with the mediator of the pending case on the mediation conflict point numbered i, respectively. i The mediation index of the mediation conflict point numbered i;

[0055] The mediation strategy generation module generates a preliminary mediation strategy based on the mediation index of each mediation conflict point in the pending case, and determines the target adjustment strategy based on the professional knowledge and experience of the mediator of the pending case;

[0056] The mediation strategy generation module obtains the mediation index of each mediation conflict point in the pending cases predicted by the mediation index prediction unit. If W i ≥0.7, it means that the mediation strategy for mediation conflict point numbered i is: support the applicant's view on mediation conflict point numbered i. If 0≤W i<0.7, it means that the mediation strategy for the mediation conflict point numbered i is: use the compromise method to mediate the mediation conflict point numbered i, for example: the applicant of the pending case requires the respondent of the pending case to repay the loan amount and interest within one year, and the respondent of the pending case does not allow it financially (can provide financial proof), so he requests to repay the loan interest within three years. When the compromise method is used to mediate the mediation conflict point, it is stipulated that the respondent of the pending case shall repay the loan amount and interest within two years. The compromise method is an informal mediation method, and its operation process is not step-by-step. As long as there is no violation of regulations or laws, it is not convenient to force it. The mediator of the pending case chooses whether to adjust the mediation strategy initially generated by the intelligent terminal based on professional knowledge and experience, and obtains the target mediation strategy based on the adjustment results;

[0057] When the applicant and the respondent of the pending case do not accept the target mediation strategy, the video automatic detection module analyzes the real-time multi-dimensional images of the applicant and the respondent of the pending case in the trial through the fusion host terminal, and predicts the shutdown time of the digital trial editing and live broadcast server through the fusion host terminal based on the analysis results, and automatically detects the video clipping time period;

[0058] The video automatic detection module includes a multi-dimensional image analysis unit, a live server shutdown time prediction unit and a video automatic detection unit;

[0059] When any one of the applicant and the respondent of the pending case does not accept the target mediation strategy, the multi-dimensional image analysis unit mediates the pending case through a trial. During the trial, the fusion host terminal collects real-time multi-dimensional images of the applicant and the respondent of the pending case. The fusion host terminal refers to a new user terminal derived from the terminal-side fusion in fixed-mobile fusion. Based on the collected real-time multi-dimensional images, the initial mediation time and the mediation end time of each mediation conflict point in the pending case are predicted. The specific method is as follows:

[0060] Based on the real-time multi-dimensional images collected by the fusion host terminal, the real-time voice and text contents of the judge, the applicant of the case to be accepted, and the respondent of the case to be accepted are collected. According to the collected real-time voice and text contents, a real-time voice and text content set of each case-related person is generated. The case-related persons include the judge, the applicant of the case to be accepted, and the respondent of the case to be accepted. The initial mediation time point and the mediation end time point of each mediation conflict point in the case to be accepted are analyzed. The specific analysis method is: in each generated real-time voice and text content set, the time when the text content of the mediation conflict point numbered i appears in each voice and text content set, and the total number of times the text content of the mediation conflict point numbered i appears are collected, and according to For Sij Predict the probability value of the initial mediation time for the mediation conflict point numbered i;

[0061] Where c=1,2,…,C, represents the numbering of the number of times each mediation conflict point appears in the generated real-time voice text content set in chronological order, C represents the total number of times each mediation conflict point appears in the generated real-time voice text content set, j=1,2,…,C and j≠c, S ic represents the time when the text content of the mediation conflict point numbered i appears for the cth time in the generated real-time voice text content set, S ij represents the time when the text content of the mediation conflict point numbered i appears for the jth time in the generated real-time voice text content set, F 1ijc 、F 2ijc 、F 3ijc Respectively expressed in [S ij ,S ic ] the matching degree between the real-time voice text content set of the judge, the applicant of the pending case, and the respondent of the pending case generated within the time period and the mediation conflict point numbered i (the matching degree between the real-time voice text content set and the mediation conflict point is calculated according to the calculation method of the matching degree between the search content and the title content, the real-time voice text content set is equivalent to the search content, the mediation conflict point is equivalent to the title content, and the calculation method of the matching degree between the search content and the title content belongs to the prior art);

[0062] like , then it means S ij is the initial mediation time of the mediation conflict point numbered i, if , then it means S ij It is not the initial mediation time of the mediation conflict point numbered i, X represents a constant and 0.5≤X≤0.7;

[0063] when When S ij The mediation end time of the mediation conflict point numbered i;

[0064] The live broadcast server shutdown time prediction unit predicts the shutdown time of the live broadcast server according to the adjustment end time of each mediation conflict point. The specific method is as follows:

[0065] In the [d, f] time period, the mediation end time of the mediation conflict point numbered i is obtained. If the total number of mediation end times obtained is less than m, it means that time f is not the shutdown time of the live broadcast server. If the total number of mediation end times obtained is equal to m, it means that time f+g is the shutdown time of the live broadcast server, where d represents the startup time of the live broadcast server, and g represents the average closing time of the trial obtained based on historical data. During the mediation process, the judge mediates each mediation conflict point in a separate mediation manner, that is, after the mediation of one mediation conflict point is completed, the next mediation conflict point will be mediated. Therefore, when the total number of mediation end times obtained is equal to the total number of mediation conflict points in the pending case, it is considered that the trial is closed and the live broadcast server is closed when the trial is closed. When the total number of mediation end times obtained is less than the total number of mediation conflict points in the pending case, it is considered that the trial is not closed and the live broadcast server is not closed at this time.

[0066] The video automatic detection unit automatically detects each cropping time period of the video uploaded to the cloud by the live broadcast server according to the prediction result of the multi-dimensional image analysis unit, and automatically crops the video uploaded to the cloud by the live broadcast server according to the detection result;

[0067] The specific method for the automatic video detection unit to automatically detect each cropping time period of the video uploaded to the cloud by the live broadcast server is:

[0068] The initial mediation time and mediation end time of each mediation conflict point are obtained to obtain the mediation time period of each mediation conflict point. Within the [d, f] time period, the video recordings uploaded to the cloud by the live broadcast server and within the time period corresponding to each mediation time period are automatically cropped;

[0069] The mediation strategy determination module is used to determine the mediation strategy for the case to be accepted based on the automatic video content of the digital trial editing and live broadcast server;

[0070] The mediation strategy determination module obtains the mediation end time of each mediation conflict point, determines the maximum value of the obtained mediation end time, and records the determined maximum mediation end time as R. The video uploaded to the cloud by the live broadcast server within the time period corresponding to the [R,f+g] time period is automatically cropped, and the voice text content in the cropped video is used as the final mediation strategy for the case to be accepted.

[0071] An automatic detection and recording method based on the integration of host terminal AI computing power includes:

[0072] S10: The case content uploaded by the applicant of the pending case to the case processing terminal is obtained, and the intelligent terminal analyzes the litigation request and evidence content in the obtained case content to generate a mediation idea. The mediator of the pending case communicates with the applicant and the respondent of the pending case based on the generated mediation idea, obtains the target integration content based on the communication results, and predicts the mediation index of each mediation conflict point in the pending case;

[0073] S20: Based on the mediation index of each mediation conflict point in the pending cases, a preliminary mediation strategy is generated, and the target adjustment strategy is determined in combination with the professional knowledge and experience of the mediators of the pending cases;

[0074] S30: When the applicant of the pending case and the respondent of the pending case do not accept the target mediation strategy, the real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case in the trial are analyzed through the fusion host terminal, and according to the analysis results, the shutdown time of the digital trial live broadcast server is predicted through the fusion host terminal, and the video clipping time period is automatically detected;

[0075] S40: Determine the mediation strategy for the case to be accepted based on the automatic video content of the digital court trial live broadcast server.

[0076] Example 1: Assume that the number of judicial interpretations associated with the mediation conflict point numbered 1 that are the same as the view supported by the applicant of the pending case is u1=5, the total number of judicial interpretations associated with the mediation conflict point numbered 1 is U1=8, and the communication time between the mediator of the pending case and the applicant of the pending case, and between the mediator of the pending case and the respondent of the pending case on the mediation conflict point numbered 1 is T respectively. 11 = 20 minutes, T 21 = 15 minutes. The time for the applicant and the respondent of the pending case to support their own opinions in the process of communicating with the mediator of the pending case on the mediation conflict point No. 1 is t respectively. 11 = 15 minutes, t 21 =15 minutes, e=2.72, then the mediation index of the mediation conflict point numbered 1 is:

[0077] ;

[0078] The mediation index of the mediation conflict point numbered 1 is 0.12.

[0079] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An automatic detection and recording system based on the AI ​​computing power of the integrated host terminal, characterized by: The system includes a case acceptance communication analysis module, a mediation strategy generation module, a video automatic detection module and a mediation strategy determination module; The communication and analysis module for accepted cases is used to obtain the case content uploaded by the applicant of the accepted case to the case processing terminal. The intelligent terminal analyzes the litigation request and evidence content in the acquired case content and generates mediation ideas. The mediator of the accepted case communicates with the applicant and the respondent of the accepted case according to the generated mediation ideas. Based on the communication results, the target integration content is obtained, and the mediation index of each mediation conflict point in the accepted case is predicted; The mediation strategy generation module preliminarily generates a mediation strategy based on the mediation index of each mediation conflict point in the pending case, and determines the target adjustment strategy in combination with the professional knowledge and experience of the mediator of the pending case; When the applicant of the pending case and the respondent of the pending case do not accept the target mediation strategy, the automatic video detection module analyzes the real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case in the trial through the fusion host terminal, and predicts the shutdown time of the digital trial live broadcast server through the fusion host terminal according to the analysis results, and automatically detects the video clipping time period; The mediation strategy determination module is used to determine the mediation strategy for the case to be accepted based on the automatic video recording content of the digital court trial editing and live broadcast server.

2. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 1 is characterized in that: The case communication analysis module includes a mediation idea generation unit, an integration screening unit and a mediation index prediction unit; The mediation idea generation unit obtains the case content uploaded by the applicant of the case to be accepted to the case processing terminal, and the intelligent terminal judges whether the litigation request or evidence content in the case content complies with the regulations, and generates a mediation idea based on the judgment result; The integration screening unit receives the mediation ideas generated by the mediation idea generation unit, collects the communication content between the mediator of the case to be accepted and the applicant and the respondent of the case to be accepted based on the received mediation ideas, and screens the case content in combination with the litigation request and evidence content in the case content. The specific screening and processing method is: if the supplementary content B of the applicant and the respondent of the case to the evidence content A that does not comply with the regulations still does not comply with the regulations, then the evidence content A is removed from the case content of the case to be accepted; if the supplementary content B of the applicant and the respondent of the case to the evidence content A that does not comply with the regulations complies with the regulations, then the evidence content A is removed from the case content of the case to be accepted, and the supplementary content B is added to the case content of the case to be accepted to obtain the target integrated content; The mediation index prediction unit receives the target integration content transmitted by the integration screening unit. The intelligent terminal determines the set of mediation conflict points of the case to be accepted based on the received target integration content, and predicts the mediation index of each mediation conflict point based on the communication results between the mediator of the case to be accepted and the applicant and respondent of the case to be accepted.

3. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 2 is characterized in that: The specific method for the mediation idea generation unit to generate the mediation idea is: if the litigation request and the evidence content in the case content are in compliance with the regulations, the generated mediation idea is to make the respondent of the case to be accepted recognize the litigation request of the applicant of the case to be accepted; if the evidence content in the case content does not comply with the regulations, the generated mediation idea is to guide the applicant of the case to be accepted and the respondent of the case to be accepted to supplement the evidence content that does not comply with the regulations; if the litigation request in the case content does not comply with the regulations, the generated mediation idea is to guide the applicant of the case to be accepted to recognize the regulations corresponding to the litigation request.

4. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 3 is characterized by: The specific method for the mediation index prediction unit to predict the mediation index of each mediation conflict point is: The judicial interpretations related to each mediation conflict point are obtained, and the viewpoints supported by each related judicial interpretation are clarified. Combined with the communication between the mediator of the pending case and the applicant and the respondent of the pending case on each mediation conflict point, the mediation index of each mediation conflict point is predicted. The specific prediction formula is: ； Among them, i=1,2,…,m, represents the number corresponding to each mediation conflict point, m represents the total number of mediation conflict points in the pending cases, and u i represents the number of judicial interpretations associated with the mediation conflict point numbered i that are the same as the view supported by the applicant of the pending case, U i represents the total number of judicial interpretations associated with the mediation conflict point numbered i, e is a constant and e>1, T 1i , T 2i represents the communication time between the mediator of the pending case and the applicant of the pending case, and between the mediator of the pending case and the respondent of the pending case on the mediation conflict point numbered i, t 1i ,t 2i W represents the time that the applicant and the respondent of the pending case support their own opinions in the process of communicating with the mediator of the pending case on the mediation conflict point numbered i, respectively. i Represents the mediation index of the mediation conflict point numbered i.

5. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 4 is characterized in that: The mediation strategy generation module obtains the mediation index of each mediation conflict point in the pending case predicted by the mediation index prediction unit. If W i ≥0.7, it means that the mediation strategy for mediation conflict point numbered i is: support the applicant's view on mediation conflict point numbered i. If 0≤W i <0.7, it means that the mediation strategy for the mediation conflict point numbered i is: use the compromise method to mediate the mediation conflict point numbered i, and the mediator of the case to be accepted chooses whether to adjust the mediation strategy initially generated by the intelligent terminal based on professional knowledge and experience, and obtains the target mediation strategy based on the adjustment result.

6. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 5 is characterized in that: The video automatic detection module includes a multi-dimensional image analysis unit, a live broadcast server shutdown time prediction unit and a video automatic detection unit; When any one of the applicant of the pending case and the respondent of the pending case does not accept the target mediation strategy, the multi-dimensional image analysis unit mediates the pending case through a court trial. During the court trial, the fusion host terminal collects real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case, and predicts the initial mediation time and the mediation end time of each mediation conflict point in the pending case based on the collected real-time multi-dimensional images; The live broadcast server shutdown time prediction unit predicts the shutdown time of the live broadcast server according to the mediation end time of each mediation conflict point; The automatic video detection unit automatically detects each cropping time period of the video uploaded to the cloud by the live broadcast server according to the prediction result of the multi-dimensional image analysis unit, and automatically crops the video uploaded to the cloud by the live broadcast server according to the detection result.

7. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 6 is characterized in that: The specific method for the multi-dimensional image analysis unit to predict the initial mediation time and mediation end time of each mediation conflict point in the accepted case is: Based on the real-time multi-dimensional images collected by the fusion host terminal, the real-time voice and text contents of the judge, the applicant of the case to be accepted, and the applicant of the case to be accepted are collected. According to the collected real-time voice and text contents, a real-time voice and text content set of each case-related person is generated, and the initial mediation time point and the mediation end time point of each mediation conflict point in the case to be accepted are analyzed. The specific analysis method is: in each generated real-time voice and text content set, the time when the text content of the mediation conflict point numbered i appears in each voice and text content set, and the total number of times the text content of the mediation conflict point numbered i appears are collected, and according to For S ij Predict the probability value of the initial mediation time for the mediation conflict point numbered i; Where c=1,2,…,C, represents the numbering of the number of times each mediation conflict point appears in the generated real-time voice text content set in chronological order, C represents the total number of times each mediation conflict point appears in the generated real-time voice text content set, j=1,2,…,C and j≠c, S ic represents the time when the text content of the mediation conflict point numbered i appears for the cth time in the generated real-time voice text content set, S ij represents the time when the text content of the mediation conflict point numbered i appears for the jth time in the generated real-time voice text content set, F 1ijc 、F 2ijc 、F 3ijc Respectively expressed in [S ij ,S ic ] The matching degree between the real-time voice text content set of the judge, the applicant of the pending case, and the respondent of the pending case generated within the time period and the mediation conflict point numbered i; like , then it means S ij is the initial mediation time of the mediation conflict point numbered i, if , then it means S ij It is not the initial mediation time of the mediation conflict point numbered i, X represents a constant and 0.5≤X≤0.7; when When S ij The mediation end time of the mediation conflict point numbered i.

8. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 7 is characterized in that: The specific method for the live broadcast service shutdown time prediction unit to predict the shutdown time of the live broadcast server is: In the time period [d, f], the mediation end time of the mediation conflict point numbered i is obtained. If the total number of mediation end times obtained is less than m, it means that time f is not the shutdown time of the live broadcast server. If the total number of mediation end times obtained is equal to m, it means that time f+g is the shutdown time of the live broadcast server, where d represents the startup time of the live broadcast server, and g represents the average closing time of the trial obtained based on historical data; The specific method of the automatic video detection unit automatically detecting each cropping time period of the video uploaded to the cloud by the live broadcast server is: The initial mediation time and mediation end time of each mediation conflict point are obtained to obtain the mediation time period of each mediation conflict point. Within the [d, f] time period, the video corresponding to each mediation time period in the video uploaded to the cloud by the live broadcast server is automatically cropped.

9. The automatic detection and recording system based on the fusion of host terminal AI computing power according to claim 8 is characterized in that: The mediation strategy determination module obtains the mediation end time of each mediation conflict point, determines the maximum value of the obtained mediation end time, and records the determined maximum value of the mediation end time as R. The video uploaded to the cloud by the live broadcast server within the time period corresponding to the [R, f+g] time period is automatically cropped, and the voice text content in the cropped video is used as the final mediation strategy for the case to be accepted.

10. An automatic detection and recording method based on the fusion of host terminal AI computing power, applied to the automatic detection and recording system based on the fusion of host terminal AI computing power according to any one of claims 1 to 9, characterized in that: The method comprises: S10: The case content uploaded by the applicant of the pending case to the case processing terminal is obtained, and the intelligent terminal analyzes the litigation request and evidence content in the obtained case content to generate a mediation idea. The mediator of the pending case communicates with the applicant and the respondent of the pending case based on the generated mediation idea, obtains the target integration content based on the communication results, and predicts the mediation index of each mediation conflict point in the pending case; S20: Based on the mediation index of each mediation conflict point in the pending cases, a preliminary mediation strategy is generated, and the target adjustment strategy is determined in combination with the professional knowledge and experience of the mediators of the pending cases; S30: When the applicant of the pending case and the respondent of the pending case do not accept the target mediation strategy, the real-time multi-dimensional images of the applicant of the pending case and the respondent of the pending case in the trial are analyzed through the fusion host terminal, and according to the analysis results, the shutdown time of the digital trial live broadcast server is predicted through the fusion host terminal, and the video clipping time period is automatically detected; S40: Determine the mediation strategy for the case to be accepted based on the automatic video content of the digital court trial live broadcast server.

Images