CN-121985150-A - Studio time delay system

CN121985150ACN 121985150 ACN121985150 ACN 121985150ACN-121985150-A

Abstract

The invention relates to the technical field of intelligent studio safe broadcasting, in particular to a studio delay system which comprises a multi-mode data acquisition module, a risk and benefit quantitative evaluation module, a dynamic delay instruction generation module, a command execution monitoring and optimization module, a real-time monitoring and adjustment process and evaluation effect after completion, and risk weight optimization according to a validity coefficient, wherein the multi-mode data acquisition module synchronously acquires audio and video data, heart rate, network and interaction data in a studio under a unified time reference, the audio and video data, the heart rate and the interaction data are packaged into synchronous data frames with time stamps, the risk and benefit quantitative evaluation module calculates a comprehensive risk index based on content, personnel states and system risk quantitative values, calculates participation degree benefit indexes by combining interaction rate and emotion analysis, the dynamic delay instruction generation module determines target delay time according to the risk and benefit indexes, and generates a control instruction after smooth transition, the instruction execution monitoring and optimization module monitors and adjusts the adjustment process in real time and evaluates the effect according to the validity coefficient, the dynamic self-adaption adjustment of delay is realized, live broadcasting safety, real-time and interaction experience are improved, and intelligence and reliability of the system are enhanced through a closed-loop optimization mechanism.

Inventors

Luo Lieyi
LOU JIAXIN
LU XUEJIE
FANG HUI
YING CHENLIANG
FENG YAOLIANG
WANG XIAOBIN
XIE JINTAO

Assignees

浙江新蓝网络传媒有限公司

Dates

Publication Date: 20260505
Application Date: 20260407

Claims (10)

1.A studio delay system, comprising: The multi-mode data acquisition module is used for synchronously acquiring multi-mode data corresponding to the studio to be evaluated and generating synchronous data frames corresponding to the multi-mode data; The multi-mode data comprises audio and video signals of the performance, personnel data, network quality data and interactive feedback data; The risk and benefit quantitative evaluation module is used for calculating a comprehensive risk index and a participation benefit index corresponding to the synchronous data frame; Combining the total number of the received interactive messages in the fixed time window with the fixed window duration, and calculating the average rate of the corresponding interactive messages; Based on the interaction file in the fixed time window corresponding to the real-time NLP emotion analysis model, calculating a net positive emotion value, and calculating to obtain a participation degree profit index corresponding to the synchronous data frame by combining the set parameter interaction rate and the emotion scaling factor; the dynamic delay instruction generation module is used for carrying out delay decision calculation and evaluating whether delay time length adjustment is needed or not, and if the delay time length adjustment is needed, an executable control instruction is generated; The instruction execution monitoring and optimizing module is used for monitoring the process of executing the control instruction in real time and evaluating the corresponding delay time length adjusting effect.
2. The studio delay system of claim 1, wherein said synchronously collecting corresponding multi-modal data in a studio to be evaluated comprises the following steps: Corresponding real-time audio and video signals, personnel data, network quality data and interactive feedback data during performance are collected; marking the acquired multi-mode data with a current time stamp, namely acquiring absolute time, and packaging the acquired multi-mode data into a standardized data packet; Grouping all standardized data packets within a set fixed time window into one synchronous data frame; For each synchronous data frame, marking the acquisition absolute time of the multi-mode data acquired in real time by the corresponding acquisition absolute time of each synchronous data frame, packaging various multi-mode data marked as the acquisition absolute time into standardized data packets, and grouping all the standardized data packets within a set fixed time window into one synchronous data frame; And subtracting the current actual time delay time length from the acquisition absolute time corresponding to each synchronous data frame based on the current actual time delay time length for each synchronous data frame to obtain a broadcasting time reference corresponding to each synchronous data frame.
3. The studio delay system of claim 2, wherein said calculating said composite risk index corresponding to said synchronization data frame comprises the steps of: And respectively calculating a content risk item quantized value, a personnel state risk item quantized value and a system risk item quantized value corresponding to each synchronous data frame, and combining preset weight factors corresponding to the content risk item quantized value, the personnel state risk item quantized value and the system risk item quantized value, and carrying out weighted calculation to obtain a comprehensive risk index corresponding to each synchronous data frame.
4. A studio delay system according to claim 3, wherein said calculating the content risk item quantization value, the personnel status risk item quantization value, and the system risk item quantization value for each synchronization data frame respectively comprises the following steps: Calculating an audio amplitude normalized burst value by combining the short-time energy of the current audio signal and the energy average value of the corresponding dynamic background; combining the motion vector of each pixel between the current frame and the previous frame to calculate a scene shear strength value; combining the audio amplitude normalized burst value, the scene shear strength value and the parameter value corresponding to the high-risk keyword frequency, and calculating to obtain a content risk item quantized value corresponding to each synchronous data frame; Acquiring a current instantaneous heart rate of a current speaker and a basic heart rate of the speaker, and calculating a personnel state risk item quantized value corresponding to each synchronous data frame by combining a historical risk coefficient of the current speaker and a preset time attenuation coefficient; and acquiring the packet loss rate and the jitter amplitude of each channel of signal, and calculating to obtain a system risk item quantized value corresponding to each synchronous data frame by combining the set maximum allowable threshold of the packet loss rate, the set maximum allowable threshold of the jitter amplitude and the weight factors corresponding to each channel of signal.
5. The studio delay system of claim 4, wherein said performing delay decision calculations comprises: and calculating a target delay time length corresponding to the synchronous data frame based on the comprehensive risk index and the participation gain index corresponding to each synchronous data frame and combining the basic safety delay time length, and a preset risk gain coefficient and a preset gain coefficient.
6. The studio delay system of claim 5, wherein said evaluating whether a delay time adjustment is required comprises: Calculating a target delay variation by combining the actual delay time corresponding to the current system and the target delay time corresponding to the synchronous data frame; by combining the target delay variation and a preset delay difference threshold, whether delay time length adjustment is needed or not is evaluated; If the delay time length adjustment is needed, the target delay time length corresponding to the synchronous data frame is constrained within a preset safety boundary.
7. The studio delay system of claim 6, wherein said generating executable control instructions comprises: calculating to obtain the instantaneous target delay time length at each moment after the delay time length adjustment is started according to the actual delay time length of the corresponding system and the final effective target delay time length when the delay time length adjustment is started and the preset total transition time length; And combining the finally effective target delay time length, the actual delay time length of the corresponding system when the delay time length starts to be adjusted, the preset total transition time length and the transition function type, and the broadcasting time reference of the next synchronous data frame, packaging the synchronous data frame into an executable control instruction, and sending the executable control instruction to the delay executing mechanism.
8. The studio delay system of claim 7, wherein said real-time monitoring of the execution of control instructions comprises: in the total transition time length of time delay time length adjustment, the actual time delay time length of the current system is combined, the comparison is carried out with the instant target time delay time length corresponding to the moment, and the adjustment deviation value of the moment is obtained through calculation; Based on the adjustment deviation value at the time, it is evaluated whether the execution process of the delay time length adjustment is faulty.
9. The studio delay system of claim 8, wherein said evaluating the corresponding delay time adjustment effect comprises: If no fault is judged to be executed, after the execution of the adjusted time delay time length, calculating a corresponding effectiveness coefficient of the time delay time length adjustment by combining the comprehensive risk indexes in the fixed time length windows before and after adjustment and the manual intervention condition of the adjustment process; And when the subsequent delay time length is adjusted, adjusting the corresponding effectiveness coefficient based on the delay time length, and adjusting the corresponding content risk item quantized value weight factor, personnel state risk item quantized value weight factor and system risk item quantized value weight factor in the risk and benefit quantized evaluation module.
10. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the system of any one of claims 1 to 9 when executing the computer program.

Description

Studio time delay system Technical Field The invention relates to the technical field of studio safe broadcasting, in particular to a studio delay system. Background The current studio live broadcast has increasingly improved requirements on safety and real-time performance, burst risks and dynamic interaction requirements are difficult to deal with by a traditional fixed delay mechanism, synchronous collection, real-time risk assessment and delay dynamic adjustment of multi-mode data are required to be realized for improving broadcasting safety and audience experience, and the system aims at realizing risk early warning and interactive gain balance in the broadcasting process through high-precision time synchronization, intelligent risk quantification and self-adaptive delay control and is suitable for live broadcast scenes such as news, variety and sports. The existing or traditional studio delay system has at least the following technical problems: 1. The traditional studio delay system lacks elaboration of synchronous acquisition and frame alignment of multi-mode data under a unified microsecond time reference, so that millisecond or even second time dislocation exists in data acquired from different devices, further, the accuracy foundation is lost when the subsequent association analysis is carried out on audio and video, physiological signals, network quality and interaction feedback, whether the disputed speaking and the sudden rise of heart rate of a speaker occur simultaneously or not cannot be judged, and therefore accurate risk assessment based on multidimensional information is difficult to realize. 2. The existing studio delay system lacks a quantitative calculation model for real-time content risks, and also lacks to calculate scene shear strength through short-time audio energy analysis and video optical flow methods and match high-risk keyword libraries through real-time voice recognition to comprehensively quantify content risk items, so that the system can only rely on subjective judgment of broadcasting guiding personnel to identify sensitive content, has lag reaction and is easy to overlook, and quantized risk indexes cannot be automatically and objectively generated at the moment of speaking violations, sudden fierce pictures or abnormal sounds, thereby missing an optimal risk prejudging window. 3. The traditional studio delay system lacks a specific method for incorporating the physiological state and the historical behavior of the personnel into a real-time risk assessment system, lacks a method for acquiring continuous speaking time through face tracking, combines the deviation of the real-time heart rate and the basic heart rate, and introduces a time attenuation coefficient based on a historical risk event to calculate personnel state risk, so that the system ignores the state change of a variable of 'people', and a host or a guest in a high risk state due to fatigue, tension or emotional agitation cannot be identified, so that early warning cannot be provided before the performer may lose the words or lose control. 4. The traditional studio delay system lacks a smooth transition control mechanism and fine monitoring on the delay adjustment process, lacks a real-time monitoring and fault judging logic on the deviation between the actual delay and the target delay in the adjustment process and also lacks a transient target delay instruction according to the total transition time length and the transition function, so that the time delay adjustment time length can possibly generate audio and video jump or asynchronous sound and picture, and is obviously perceived by a spectator. Disclosure of Invention Aiming at the defects existing in the prior art, the invention provides a studio delay system, which can effectively solve the problem that a fixed delay strategy in the background art cannot dynamically balance broadcasting safety and real-time interaction. In order to solve the technical problems, the invention adopts the following technical scheme that the invention provides a studio delay system, which comprises: the multi-mode data acquisition module is used for synchronously acquiring multi-mode data corresponding to the studio to be evaluated and generating synchronous data frames corresponding to the multi-mode data. The multi-mode data comprises audio and video signals of the performance, personnel data, network quality data and interactive feedback data. And the risk and benefit quantitative evaluation module is used for calculating the comprehensive risk index and the participation benefit index corresponding to the synchronous data frame. The dynamic delay instruction generation module is used for carrying out delay decision calculation and evaluating whether delay time length adjustment is needed or not, and if the delay time length adjustment is needed, an executable control instruction is generated. The instruction execution monitoring and optimizing module is used for