AI-based audio and video data processing method and system

Abstract

The present invention provides an AI-based audio and video data processing method and system; the AI-based audio and video data processing method includes: after determining that the target audio and video clip belongs to the high-dynamic voice characteristic type, obtaining the initial scheduling priority value generated for the target audio and video clip in the link end warning state, and obtaining the historical transmission data record of the associated device that carries the target audio and video clip transmission task. The present invention can dynamically adjust the scheduling priority value of the audio and video clip according to the real-time link performance and historical data trends in the link end warning state, significantly improving the efficiency and stability of the link transmission. The core innovation lies in the combination of a first correction parameter and a second correction parameter, wherein the first correction parameter is dynamically corrected based on the real-time link performance, and the second correction parameter is optimized and adjusted according to the trend change of historical data.

Description

Technical Field

[0001] The present invention relates to the technical field of audio and video data transmission optimization, and in particular to an AI-based audio and video data processing method and system. Background Art

[0002] With the widespread adoption of audio, video, and real-time data transmission technologies, network link quality has become a critical factor affecting transmission stability and user experience. Especially when dealing with highly dynamic audio and video content, fluctuations in link performance, such as bandwidth, latency, and packet loss, often directly impact transmission quality. Existing audio and video data transmission strategies typically rely on static priority scheduling algorithms based on metrics such as link bandwidth, latency, and packet loss rate. However, these strategies fail to fully consider the real-time alert status of link endpoint devices. In practice, a link endpoint alert is triggered when a link endpoint device (such as a user terminal or access device) detects that link performance is approaching or exceeding the lower quality of service (QoS) threshold. At this point, the system must promptly adjust its scheduling strategy to prevent degradation in audio and video transmission quality caused by declining link quality.

[0003] While existing scheduling priority allocation methods base scheduling on the warning status of devices at the end of the link, they still suffer from static settings and an inability to flexibly respond to dynamic changes in the link. In network environments with multi-hop relays, fluctuating bandwidth, or large link delays, traditional methods are unable to dynamically adjust based on actual link conditions. This can result in the transmission priority of critical content such as highly dynamic voice clips not being guaranteed in a timely manner. Therefore, existing scheduling mechanisms fail to effectively integrate the dynamic changes in link status with the characteristics of audio and video data, and are unable to accurately identify and adapt to real-time transmission requirements, affecting overall transmission efficiency and quality. Summary of the Invention

[0004] The purpose of the present invention is to provide an AI-based audio and video data processing method and system to solve the technical problem of insufficient link dynamic scheduling in the prior art.

[0005] The present invention is implemented as follows: an AI-based audio and video data processing method, the method comprising:

[0006] After determining that the target audio and video clip belongs to the high-dynamic voice characteristic type, obtaining the initial scheduling priority value generated for the target audio and video clip in the link end warning state, and obtaining the historical transmission data record of the associated device that carries the target audio and video clip transmission task;

[0007] Parsing historical transmission data records, screening out a number of first-category historical segments that are consistent with the environmental characteristics of the target audio and video segments and belong to the high-dynamic voice characteristic type, and a number of second-category historical segments that do not belong to the high-dynamic voice characteristic type;

[0008] Counting the probability of link transmission performance degradation exceeding a preset threshold in a number of first-category historical segments and a number of second-category historical segments, respectively, and determining a first correction parameter based on a deviation between the two;

[0009] Calculating a reconstruction cost value corresponding to each first-category historical segment, and determining a second correction parameter based on a change trend characteristic of the reconstruction cost value;

[0010] The initial scheduling priority value is corrected in combination with the first correction parameter and the second correction parameter.

[0011] As a further limitation of the technical solution of the embodiment of the present invention, the steps of respectively counting the probability of link transmission performance degradation exceeding a preset threshold in a plurality of first-category historical segments and a plurality of second-category historical segments, and determining the first correction parameter based on the deviation between the two include:

[0012] Sequentially parsing each of the first-category historical segments and the second-category historical segments, extracting link transmission degradation event data recorded in each historical segment to determine the corresponding link transmission performance degradation magnitude;

[0013] Count the number of segments in which the link transmission performance degradation exceeds a preset threshold in a number of first-category historical segments and a number of second-category historical segments, and calculate the corresponding occurrence probability;

[0014] A first correction parameter is determined based on a deviation between the probability of occurrence that the link transmission performance degradation range in the first type of historical segments and the second type of historical segments exceeds a preset threshold standard.

[0015] As a further limitation of the technical solution of an embodiment of the present invention, the link transmission degradation event data includes packet loss rate, delay jitter amplitude and instantaneous bandwidth fluctuation amplitude, and the link transmission performance degradation amplitude is calculated by weighted combination of packet loss rate, delay jitter amplitude and instantaneous bandwidth fluctuation amplitude.

[0016] As a further limitation of the technical solution of the embodiment of the present invention, the step of calculating the reconstruction cost value corresponding to each first-category historical segment and determining the second correction parameter according to the change trend characteristics of the reconstruction cost value includes:

[0017] Parsing the link transmission data recorded in each first-category historical segment, extracting the actual number of retransmissions and transmission delay corresponding to each first-category historical segment, and calculating the reconstruction cost value corresponding to each first-category historical segment based on the actual number of retransmissions and transmission delay;

[0018] Count the reconstruction cost values corresponding to all first-category historical segments, and construct a reconstruction cost change curve according to the time sequence of the segments;

[0019] The average slope of the reconstruction cost value change curve is calculated and used as the second correction parameter.

[0020] As a further limitation of the technical solution of the embodiment of the present invention, the actual number of retransmissions is obtained by the number of data retransmissions occurring during the link transmission process, and the transmission delay is calculated by the time difference between the data sending time and the receiving time of each first-category historical segment;

[0021] The reconstruction cost value is calculated by performing a weighted linear combination of the actual number of retransmissions and the transmission delay.

[0022] As a further limitation of the technical solution of the embodiment of the present invention, the step of correcting the initial scheduling priority value by combining the first correction parameter and the second correction parameter includes:

[0023] Retrieving a preset dispatch priority value correction formula, substituting the first correction parameter and the second correction parameter into the dispatch priority value correction formula, correcting the initial dispatch priority value, and obtaining a corrected dispatch priority value;

[0024] The modified scheduling priority value is applied to the link resource scheduling policy of the target video segment to dynamically adjust the scheduling priority level of the target video segment in the transmission link.

[0025] As a further limitation of the technical solution of the embodiment of the present invention, the scheduling priority value correction formula is:

[0026] ;

[0027] in Refers to the revised dispatch priority value;

[0028] Refers to the initial scheduling priority value, Refers to the first correction parameter, that is, the deviation of the probability of link transmission performance degradation exceeding the preset threshold standard in the first and second historical segments, Refers to the adjustment coefficient corresponding to the first correction parameter, Refers to the second correction parameter, that is, the average slope of the reconstruction cost value change curve, Refers to the adjustment coefficient corresponding to the second correction parameter;

[0029] In the scheduling priority value correction formula, ;

[0030] in Refers to the probability of link transmission performance degradation exceeding the preset threshold standard in the first type of historical segment, Refers to the probability of link transmission performance degradation exceeding the preset threshold standard in the second type of historical fragments. It is the preset minimum protection value, used to avoid calculation anomalies when the denominator is zero or close to zero.

[0031] An AI-based audio and video data processing system, the system comprising: a data acquisition module, a data screening module, a first correction parameter determination module, a second correction parameter determination module, and a correction module, wherein:

[0032] A data acquisition module is used to obtain, after determining that the target audio and video clip belongs to the high-dynamic voice characteristic type, an initial scheduling priority value generated for the target audio and video clip in the link end warning state, and obtain historical transmission data records of the associated device that carries the target audio and video clip transmission task;

[0033] A data screening module is used to parse historical transmission data records and screen out a number of first-category historical segments that are consistent with the environmental characteristics of the target audio and video segments and belong to the high-dynamic voice characteristic type, and a number of second-category historical segments that do not belong to the high-dynamic voice characteristic type;

[0034] A first correction parameter determination module is configured to calculate the probability of link transmission performance degradation exceeding a preset threshold in a plurality of first-category historical segments and a plurality of second-category historical segments, and determine a first correction parameter based on a deviation between the two.

[0035] a second correction parameter determination module, configured to calculate a reconstruction cost value corresponding to each first-category historical segment, and determine a second correction parameter based on a change trend characteristic of the reconstruction cost value;

[0036] The correction module is used to correct the initial scheduling priority value by combining the first correction parameter and the second correction parameter.

[0037] As a further limitation of the technical solution of the embodiment of the present invention, the first correction parameter determination module specifically includes:

[0038] a historical segment parsing unit, configured to sequentially parse each of the first and second historical segments, extract link transmission degradation event data recorded in each historical segment, and determine a corresponding link transmission performance degradation magnitude; the link transmission degradation event data includes packet loss rate, delay jitter magnitude, and instantaneous bandwidth fluctuation magnitude, and the link transmission performance degradation magnitude is calculated by performing a weighted combination of the packet loss rate, delay jitter magnitude, and instantaneous bandwidth fluctuation magnitude;

[0039] an occurrence probability calculation unit, configured to count the number of segments in which the link transmission performance degradation exceeds a preset threshold value in the plurality of first-category historical segments and the plurality of second-category historical segments, and calculate the corresponding occurrence probability thereof;

[0040] The deviation amplitude determining unit is used to determine a first correction parameter based on the deviation amplitude of the probability of the link transmission performance degradation amplitude exceeding a preset threshold standard in the first type of historical segments and the second type of historical segments.

[0041] As a further limitation of the technical solution of the embodiment of the present invention, the second correction parameter determination module specifically includes:

[0042] a transmission data parsing unit, configured to parse the link transmission data recorded in each first-category historical segment, extract the actual number of retransmissions and the transmission delay corresponding to each first-category historical segment, and calculate a reconstruction cost value corresponding to each first-category historical segment based on the actual number of retransmissions and the transmission delay; the actual number of retransmissions being obtained by the number of data retransmissions occurring during the link transmission process, the transmission delay being calculated by the time difference between the data transmission time and the data reception time of each first-category historical segment; and the reconstruction cost being calculated by a weighted linear combination of the actual number of retransmissions and the transmission delay;

[0043] A change curve construction unit is used to count the reconstruction cost values corresponding to all first-category historical segments and construct a reconstruction cost value change curve according to the time sequence of the segments;

[0044] The average slope calculation unit is used to calculate the average slope of the reconstruction cost value change curve and use the average slope as the second correction parameter.

[0045] By adopting the above technical solution, the present invention has the following beneficial effects:

[0046] This paper proposes an AI-based audio and video data processing method that dynamically adjusts the scheduling priority of audio and video clips based on real-time link performance and historical data trends during a link end warning state, significantly improving the efficiency and stability of link transmission. The core innovation lies in the combination of a first correction parameter (dynamically adjusted based on real-time link performance (such as packet loss rate and latency)) and a second correction parameter (optimized and adjusted based on historical data trends (such as changes in reconstruction cost).

[0047] Through this comprehensive correction mechanism, the present invention effectively addresses the existing problem of insufficient consideration of link fluctuations and long-term trend changes. Especially for highly dynamic voice clips in unstable network environments, the present invention prioritizes their transmission based on early warning signals from the link end, thereby improving the quality and reliability of audio and video transmission. Through this innovative mechanism, the present invention provides more stable and reliable support for real-time communication and remote video conferencing in complex network environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 A flowchart of a method provided by an embodiment of the present invention;

[0050] Figure 2 A flowchart of determining a first correction parameter in the method provided in an embodiment of the present invention;

[0051] Figure 3 A flowchart of determining a second correction parameter in the method provided in an embodiment of the present invention;

[0052] Figure 4 A flowchart of correcting the initial scheduling priority value in the method provided in an embodiment of the present invention;

[0053] Figure 5 An application architecture diagram of the system provided by an embodiment of the present invention;

[0054] Figure 6 A structural block diagram of a first correction parameter determination module in a system provided by an embodiment of the present invention;

[0055] Figure 7 This is a structural block diagram of the second correction parameter determination module in the system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] The present invention will be further explained below with reference to specific embodiments.

[0058] Figure 1 A flow chart of a method provided by an embodiment of the present invention is shown.

[0059] Specifically, an AI-based audio and video data processing method includes the following steps:

[0060] Step S100, after determining that the target audio and video clip belongs to the high dynamic voice feature type, obtain the initial scheduling priority value generated for the target audio and video clip in the link end warning state, and obtain the historical transmission data record of the associated device that carries the target audio and video clip transmission task.

[0061] In an embodiment of the present invention, with the widespread application of audio and video transmission technology, the quality and stability of the link have become key factors affecting the user experience. In particular, during the transmission of real-time audio and video data, fluctuations in link performance (such as packet loss, latency, bandwidth fluctuations, etc.) directly affect the transmission quality of audio and video content. Therefore, in the case of unstable links, it is particularly important to adopt a method for dynamically adjusting the scheduling priority value. The AI-based audio and video data processing method of the present invention can dynamically adjust the scheduling priority value by intelligently evaluating the characteristics of the target audio and video clips and the link status, thereby ensuring the priority transmission of high-dynamic audio and video clips and optimizing the overall transmission quality.

[0062] Link end warning status refers to a system-triggered warning state when link performance indicators (such as bandwidth, latency, and packet loss rate) approach or exceed preset warning thresholds during transmission at the end device of a link in the network (such as a user terminal or access device). Specifically, when the transmission performance monitored by the device reaches or approaches the predetermined Quality of Service (QoS) lower limit, the system triggers a warning signal, indicating that the current link transmission quality may not meet the requirements of real-time audio and video content transmission, resulting in quality degradation or transmission delays.

[0063] Highly dynamic speech refers to audio signals with significant fluctuations and variability in speech content, such as changes in speaking rate, volume, and tone. Compared to other types of audio and video clips, highly dynamic speech clips place more stringent requirements on link performance. These clips are particularly sensitive to factors such as latency, packet loss, and bandwidth fluctuations. Any minor transmission issue can significantly impact speech continuity and clarity. Therefore, for these clips, dynamic priority scheduling is crucial to ensure their transmission priority and guarantee transmission quality.

[0064] The initial scheduling priority value is derived based on the characteristics of the target audio and video clip and the current link status. This value is typically set based on the type of audio and video data, transmission conditions, and device performance. In existing technologies, scheduling priority values are typically determined through static settings or link status assessments. For example, an initial priority value is assigned to each clip using a fixed algorithm or rule based on criteria such as network bandwidth, latency, and packet loss.

[0065] Associated devices refer to devices involved in the transmission, processing, and scheduling of data streams during the transmission of target audio and video clips. Specifically, these devices include user terminals (such as smartphones, computers, and tablets), network access devices (such as routers, switches, and base stations), and transmission devices (such as video encoders, transmission servers, and CDN nodes). These devices are responsible for transmitting audio and video data streams, and their link status directly affects the transmission quality of the audio and video data.

[0066] Historical transmission data records are typically collected through network monitoring systems, link management platforms, or built-in monitoring modules on devices. These records contain information about the link's transmission performance, including but not limited to the following: packet loss rate (the percentage of packets lost per unit time); latency (the time it takes for a packet to travel from the source device to the destination); bandwidth fluctuation (the range of fluctuations in network bandwidth per unit time); retransmission count (the number of times a packet is retransmitted); throughput (the amount of data successfully transmitted per unit time); and link load (the utilization rate of the link).

[0067] Furthermore, the AI-based audio and video data processing method further includes the following steps:

[0068] Step S200 , parsing historical transmission data records, screening out a number of first-category historical segments that are consistent with the target audio and video segment environment characteristics and belong to the high dynamic voice characteristic type, and a number of second-category historical segments that do not belong to the high dynamic voice characteristic type.

[0069] In an embodiment of the present invention, the historical transmission data record includes not only conventional link performance information, but also environmental feature data used to determine whether the target audio and video clip belongs to the type of high-dynamic voice characteristics. These environmental feature data can be obtained by performing a detailed analysis of the voice signal of the target audio and video clip. Specifically, it includes spectrum analysis data, which is used to evaluate the frequency distribution and frequency change trend of the audio signal; volume fluctuation data, which is used to detect the volume change of the audio signal in the time dimension, especially when the volume fluctuation is large in the speech; and speech rate change data, which is used to measure the rate of change of speech rate in the speech clip. Rapid changes in speech rate are usually associated with high-dynamic voice characteristic clips.

[0070] Through comprehensive analysis of these environmental feature data, the system can determine whether the target audio and video clip belongs to the high-dynamic voice characteristic type. Next, the system selects several historical clips that are consistent with the environmental characteristics of the target clip from the historical transmission data records. The screening process uses a similarity calculation method, for example, by calculating the similarity between the target clip and the historical clip in terms of spectral characteristics, volume fluctuations, and speech speed changes, to identify historical clips that meet the conditions. The system further classifies these clips, first screening out the first category of historical clips that meet the characteristics of the target clip and belong to the high-dynamic voice characteristic type, and then screening out the second category of historical clips that do not have the high-dynamic voice characteristic type of the target clip.

[0071] This process achieves accurate screening of historical clips through dynamic feature matching, similarity calculation and classification tagging, enabling the system to provide more accurate transmission scheduling optimization solutions based on the environmental characteristics and dynamic voice characteristics of the target audio and video clips.

[0072] Furthermore, the AI-based audio and video data processing method further includes the following steps:

[0073] Step S300 , counting the probability of link transmission performance degradation exceeding a preset threshold in a plurality of first-category historical segments and a plurality of second-category historical segments, and determining a first correction parameter based on the deviation between the two.

[0074] Specifically, Figure 2 A flow chart for determining a first correction parameter is shown.

[0075] The process of counting the probability of link transmission performance degradation exceeding a preset threshold value in a plurality of first-category historical segments and a plurality of second-category historical segments, and determining the first correction parameter based on the deviation between the two, specifically includes the following steps:

[0076] Step S301: parsing each of the first-category historical segments and the second-category historical segments in sequence, extracting link transmission degradation event data recorded in each historical segment to determine the corresponding link transmission performance degradation magnitude;

[0077] Step S302: Count the number of segments in which the link transmission performance degradation exceeds a preset threshold in the first category historical segments and the second category historical segments, and calculate the corresponding occurrence probability.

[0078] Step S303 : determining a first correction parameter based on the deviation of the probability of link transmission performance degradation exceeding a preset threshold in the first type of historical segments and the second type of historical segments.

[0079] The link transmission degradation event data includes packet loss rate, delay jitter amplitude and instantaneous bandwidth fluctuation amplitude. The link transmission performance degradation amplitude is calculated by weighted combination of packet loss rate, delay jitter amplitude and instantaneous bandwidth fluctuation amplitude.

[0080] In the embodiment of the present invention, the specific implementation process of step S302 is as follows:

[0081] First, the system analyzes each of the first and second historical segments in turn, extracting the link transmission performance degradation event data within each segment. This data includes packet loss rate, delay jitter amplitude, and instantaneous bandwidth fluctuation amplitude. By performing a weighted combination calculation on this data, the system determines the link transmission performance degradation amplitude for each segment.

[0082] The system then counts all historical segments of the first and second categories and calculates the number of segments in which link transmission performance degradation exceeds a preset threshold. Next, the system calculates the probability of occurrence of segments exceeding the threshold within each segment category. Specifically, the system divides the number of segments in each segment category in which link transmission performance degradation exceeds the threshold by the total number of segments in that category to determine the probability of occurrence.

[0083] Next, the system determines the first correction parameter based on the difference between the probability of link transmission performance degradation exceeding the threshold for the first and second historical segments. This difference can be calculated by calculating the deviation between the two probabilities. If the probability of link transmission performance degradation exceeding the threshold for the first historical segment is greater than that for the second historical segment, the deviation is large, indicating that the first segment (i.e., the high-dynamic voice segment) is more likely to experience performance degradation during link transmission, and the system needs to adjust its priority accordingly. If the difference between the two probabilities is small, the link performance of the high-dynamic voice segment is not much different from that of the normal segment, and the corresponding priority adjustment will be smaller.

[0084] The significance and technical benefit of this process lies in that by comparing the probability of transmission performance degradation between the first and second categories of historical segments, the system can more accurately measure the performance differences between high-dynamic voice segments and other segments during link transmission. This allows the system to dynamically adjust the scheduling priority of these segments based on actual conditions, ensuring that high-dynamic voice segments receive priority when link conditions are poor, thereby improving audio and video transmission quality and user experience.

[0085] Furthermore, the AI-based audio and video data processing method further includes the following steps:

[0086] Step S400 : calculating the reconstruction cost value corresponding to each first-category historical segment, and determining a second correction parameter according to a change trend characteristic of the reconstruction cost value.

[0087] Specifically, Figure 3 A flow chart for determining the second correction parameter is shown.

[0088] Calculating the reconstruction cost value corresponding to each first-category historical segment and determining the second correction parameter according to the change trend characteristics of the reconstruction cost value specifically includes the following steps:

[0089] Step S401: parsing the link transmission data recorded in each first-category historical segment, extracting the actual number of retransmissions and the transmission delay corresponding to each first-category historical segment, and calculating the reconstruction cost value corresponding to each first-category historical segment based on the actual number of retransmissions and the transmission delay;

[0090] Step S402: Count the reconstruction cost values corresponding to all first-category historical segments, and construct a reconstruction cost value change curve according to the segment time sequence;

[0091] Step S403 : Calculate the average slope of the reconstruction cost value change curve, and use the average slope as the second correction parameter.

[0092] The actual number of retransmissions is obtained by the number of data retransmissions occurring during the link transmission process, and the transmission delay is calculated by the time difference between the data sending time and the receiving time of each first-category historical fragment; the reconstruction cost value is calculated by a weighted linear combination of the actual number of retransmissions and the transmission delay.

[0093] In the embodiment of the present invention, the specific implementation process of step S401 is as follows:

[0094] First, the system parses the link transmission data recorded in each first-category historical segment and extracts the actual number of retransmissions and transmission delay corresponding to each segment. The actual number of retransmissions is obtained by monitoring the number of packet retransmissions during link transmission, that is, calculating the retransmission frequency of each segment. Transmission delay is obtained by calculating the time difference between the data transmission time and the reception time of each historical segment. It is usually expressed as the time required for the packet to propagate across the network.

[0095] Next, the system calculates the reconstruction cost for each first-category historical segment by performing a weighted linear combination of the actual number of retransmissions and the transmission delay. Before calculating the reconstruction cost for each first-category historical segment, the system first normalizes the number of retransmissions and the transmission delay. Because the number of retransmissions and the transmission delay have different dimensions and magnitudes, a direct weighted linear combination could result in one metric overly influential in the final result. Therefore, normalization is required to ensure that both metrics are weighted equally.

[0096] After normalization, the system weights the number of retransmissions and transmission delay according to a preset weighting factor and sums them to obtain the reconstruction cost for each segment. This weighted linear combination allows the system to comprehensively consider the impact of link stability (reflected by the number of retransmissions) and transmission delay (reflected by the transmission delay) on the reconstruction cost, thereby calculating the reconstruction difficulty and resource consumption required for each segment.

[0097] In step S402, the system calculates the reconstruction cost values for all first-category historical segments and constructs a reconstruction cost change curve based on the chronological order of the segments. Specifically, the system arranges the reconstruction cost values for each historical segment in chronological order, resulting in a curve that reflects the change in reconstruction cost over time. This curve can demonstrate the fluctuations in reconstruction demand within each time period during the transmission process.

[0098] In step S403, the system calculates the average slope of the reconstruction cost curve and uses this average slope as the second correction parameter. By calculating the average slope of the curve, the system can determine the overall trend of the reconstruction cost over time. If the slope of the curve is positive, it indicates that the reconstruction cost value is gradually increasing over time, indicating that the quality of the transmission link may be deteriorating, resulting in a higher reconstruction cost. If the slope of the curve is negative, it indicates that the reconstruction cost value is gradually decreasing, which may mean that the link quality is improving and the reconstruction cost is decreasing. This average slope value, as the second correction parameter, can reflect the volatility of the link during the transmission process and its impact on the reconstruction cost.

[0099] The significance of calculating the average slope of the reconstruction cost curve and using it as the second correction parameter lies in its ability to effectively reflect the changing trend of link transmission quality. This method not only allows the system to understand the current link status in real time but also predict future link changes based on historical data. This trend-based adjustment parameter helps the system adjust priorities in real time within dynamic network environments, ensuring that audio and video clips that may require more resources for reconstruction during transmission are prioritized, thereby optimizing the efficiency and quality of the entire transmission process.

[0100] Furthermore, the AI-based audio and video data processing method further includes the following steps:

[0101] Step S500: Correct the initial scheduling priority value by combining the first correction parameter and the second correction parameter.

[0102] Specifically, Figure 4 A flow chart for modifying the initial scheduling priority value is shown.

[0103] The correction of the initial scheduling priority value by combining the first correction parameter and the second correction parameter specifically includes the following steps:

[0104] Step S501: Retrieve a preset scheduling priority value correction formula, substitute the first correction parameter and the second correction parameter into the scheduling priority value correction formula, correct the initial scheduling priority value, and obtain a corrected scheduling priority value;

[0105] Step S502: Apply the modified scheduling priority value to the link resource scheduling policy of the target video segment to dynamically adjust the scheduling priority level of the target video segment in the transmission link.

[0106] The scheduling priority value correction formula is:

[0107] ;

[0108] in Refers to the revised scheduling priority value, Refers to the initial scheduling priority value, Refers to the first correction parameter, that is, the deviation of the probability of link transmission performance degradation exceeding the preset threshold standard in the first and second historical segments, Refers to the adjustment coefficient corresponding to the first correction parameter, Refers to the second correction parameter, that is, the average slope of the reconstruction cost value change curve, Refers to the adjustment coefficient corresponding to the second correction parameter;

[0109] In the scheduling priority value correction formula, ,in Refers to the probability of link transmission performance degradation exceeding the preset threshold standard in the first type of historical segment, Refers to the probability of link transmission performance degradation exceeding the preset threshold standard in the second type of historical fragments. It is the preset minimum protection value, used to avoid calculation anomalies when the denominator is zero or close to zero.

[0110] In this embodiment of the present invention, the purpose of combining the first and second correction parameters for comprehensive correction is to comprehensively adjust the scheduling priority of the target audio and video clips to cope with changes in the network link and the transmission requirements of different types of audio and video clips. The first correction parameter is primarily based on the real-time status of the link transmission performance, such as factors such as packet loss rate and latency. It reflects the quality of the current link and can adjust the priority of the clip in a timely manner, especially when the network fluctuates significantly during transmission. The second correction parameter is based on trends in historical data, such as the changing trend of the reconstruction cost value, which captures the fluctuation of the network link over time.

[0111] This comprehensive correction allows the system to more flexibly respond to sudden link fluctuations and long-term trends based on historical data, enabling more accurate adjustment of transmission priorities and ensuring that highly dynamic voice clips are prioritized even in unstable network environments. The combination of the first and second correction parameters provides a solution that balances real-time and historical factors, avoiding the potential bias caused by a single factor. This allows the system to more accurately determine the transmission priority of each audio and video clip, improving overall transmission efficiency and stability.

[0112] For example, suppose the system is transmitting a highly dynamic voice segment. First, the system calculates a first correction parameter based on real-time link conditions. If the link bandwidth fluctuates significantly, the packet loss rate is high, or the latency increases, the system will increase the priority of the segment to ensure its transmission. Next, the system calculates a second correction parameter by analyzing historical data, particularly trends in reconstruction costs. If historical data indicates that reconstruction costs are increasing over time, the system will also increase the priority of the segment to ensure smooth transmission.

[0113] Ultimately, the revised scheduling priority value is calculated by incorporating these two modified parameters into the initial priority value. This allows the system to dynamically adjust scheduling priorities based on real-time link status and historical transmission data, ensuring that highly dynamic voice clips are prioritized even in poor network conditions, thereby improving overall audio and video transmission quality.

[0114] The scheduling priority value correction formula proposed in this paper is only a straightforward and effective solution, and it can be further optimized through various other calculation methods. For example, in some systems, the correction coefficient can be adjusted according to different scenarios or real-time link conditions, making the correction process more flexible and precise. By incorporating technologies such as machine learning, the system can also adaptively adjust these correction coefficients based on historical data, achieving even higher optimization results.

[0115] Further, Figure 5 The application architecture diagram of the system provided by the embodiment of the present invention is shown.

[0116] In another preferred embodiment of the present invention, an AI-based audio and video data processing system includes:

[0117] The data acquisition module 100 is used to obtain the initial scheduling priority value generated for the target audio and video clip in the link end warning state after determining that the target audio and video clip belongs to the high-dynamic voice feature type, and obtain the historical transmission data record of the associated device that carries the target audio and video clip transmission task.

[0118] In an embodiment of the present invention, with the widespread application of audio and video transmission technology, the quality and stability of the link have become key factors affecting the user experience. In particular, during the transmission of real-time audio and video data, fluctuations in link performance (such as packet loss, latency, bandwidth fluctuations, etc.) directly affect the transmission quality of audio and video content. Therefore, in the case of unstable links, it is particularly important to adopt a method for dynamically adjusting the scheduling priority value. The AI-based audio and video data processing method of the present invention can dynamically adjust the scheduling priority value by intelligently evaluating the characteristics of the target audio and video clips and the link status, thereby ensuring the priority transmission of high-dynamic audio and video clips and optimizing the overall transmission quality.

[0119] Link end warning status refers to a system-triggered warning state when link performance indicators (such as bandwidth, latency, and packet loss rate) approach or exceed preset warning thresholds during transmission at the end device of a link in the network (such as a user terminal or access device). Specifically, when the transmission performance monitored by the device reaches or approaches the predetermined Quality of Service (QoS) lower limit, the system triggers a warning signal, indicating that the current link transmission quality may not meet the requirements of real-time audio and video content transmission, resulting in quality degradation or transmission delays.

[0120] Highly dynamic speech refers to audio signals with significant fluctuations and variability in speech content, such as changes in speaking rate, volume, and tone. Compared to other types of audio and video clips, highly dynamic speech clips place more stringent requirements on link performance. These clips are particularly sensitive to factors such as latency, packet loss, and bandwidth fluctuations. Any minor transmission issue can significantly impact speech continuity and clarity. Therefore, for these clips, dynamic priority scheduling is crucial to ensure their transmission priority and guarantee transmission quality.

[0121] The initial scheduling priority value is derived based on the characteristics of the target audio and video clip and the current link status. This value is typically set based on the type of audio and video data, transmission conditions, and device performance. In existing technologies, scheduling priority values are typically determined through static settings or link status assessments. For example, an initial priority value is assigned to each clip using a fixed algorithm or rule based on criteria such as network bandwidth, latency, and packet loss.

[0122] Associated devices refer to devices involved in the transmission, processing, and scheduling of data streams during the transmission of target audio and video clips. Specifically, these devices include user terminals (such as smartphones, computers, and tablets), network access devices (such as routers, switches, and base stations), and transmission devices (such as video encoders, transmission servers, and CDN nodes). These devices are responsible for transmitting audio and video data streams, and their link status directly affects the transmission quality of the audio and video data.

[0123] Furthermore, the AI-based audio and video data processing system also includes:

[0124] The data screening module 200 is used to parse the historical transmission data records and screen out several first-category historical segments that are consistent with the target audio and video segment environment characteristics and belong to the high dynamic voice characteristic type, and several second-category historical segments that do not belong to the high dynamic voice characteristic type.

[0125] In an embodiment of the present invention, the historical transmission data record includes not only conventional link performance information, but also environmental feature data used to determine whether the target audio and video clip belongs to the type of high-dynamic voice characteristics. These environmental feature data can be obtained by performing a detailed analysis of the voice signal of the target audio and video clip. Specifically, it includes spectrum analysis data, which is used to evaluate the frequency distribution and frequency change trend of the audio signal; volume fluctuation data, which is used to detect the volume change of the audio signal in the time dimension, especially when the volume fluctuation is large in the speech; and speech rate change data, which is used to measure the rate of change of speech rate in the speech clip. Rapid changes in speech rate are usually associated with high-dynamic voice characteristic clips.

[0126] Through comprehensive analysis of these environmental feature data, the system can determine whether the target audio and video clip belongs to the high-dynamic voice characteristic type. Next, the system selects several historical clips that are consistent with the environmental characteristics of the target clip from the historical transmission data records. The screening process uses a similarity calculation method, for example, by calculating the similarity between the target clip and the historical clip in terms of spectral characteristics, volume fluctuations, and speech speed changes, to identify historical clips that meet the conditions. The system further classifies these clips, first screening out the first category of historical clips that meet the characteristics of the target clip and belong to the high-dynamic voice characteristic type, and then screening out the second category of historical clips that do not have the high-dynamic voice characteristic type of the target clip.

[0127] This process achieves accurate screening of historical clips through dynamic feature matching, similarity calculation and classification tagging, enabling the system to provide more accurate transmission scheduling optimization solutions based on the environmental characteristics and dynamic voice characteristics of the target audio and video clips.

[0128] Furthermore, the AI-based audio and video data processing system also includes:

[0129] The first correction parameter determination module 300 is used to count the probability of link transmission performance degradation exceeding a preset threshold standard in a number of first-type historical segments and a number of second-type historical segments, and determine a first correction parameter based on the deviation between the two.

[0130] Specifically, Figure 6FIG. 3 is a structural block diagram of a first correction parameter determination module 300 in a system provided by an embodiment of the present invention.

[0131] In a preferred embodiment of the present invention, the first correction parameter determination module 300 specifically includes:

[0132] The historical segment parsing unit 301 is configured to sequentially parse each of the first and second historical segments, extract the link transmission degradation event data recorded in each historical segment, and determine the corresponding link transmission performance degradation magnitude. The link transmission degradation event data includes packet loss rate, delay jitter magnitude, and instantaneous bandwidth fluctuation magnitude. The link transmission performance degradation magnitude is calculated by performing a weighted combination of the packet loss rate, delay jitter magnitude, and instantaneous bandwidth fluctuation magnitude.

[0133] An occurrence probability calculation unit 302 is configured to count the number of segments in which the link transmission performance degradation exceeds a preset threshold in the plurality of first-category historical segments and the plurality of second-category historical segments, and calculate the corresponding occurrence probability of each segment;

[0134] The deviation amplitude determining unit 303 is configured to determine a first correction parameter based on the deviation amplitude of the occurrence probability that the link transmission performance degradation amplitude in the first type of historical segments and the second type of historical segments exceeds a preset threshold standard.

[0135] In the embodiment of the present invention, the specific working process of the historical segment parsing unit 301 is as follows:

[0136] First, the system analyzes each of the first and second historical segments in turn, extracting the link transmission performance degradation event data within each segment. This data includes packet loss rate, delay jitter amplitude, and instantaneous bandwidth fluctuation amplitude. By performing a weighted combination calculation on this data, the system determines the link transmission performance degradation amplitude for each segment.

[0137] The system then counts all historical segments of the first and second categories and calculates the number of segments in which link transmission performance degradation exceeds a preset threshold. Next, the system calculates the probability of occurrence of segments exceeding the threshold within each segment category. Specifically, the system divides the number of segments in each segment category in which link transmission performance degradation exceeds the threshold by the total number of segments in that category to determine the probability of occurrence.

[0138] Next, the system determines the first correction parameter based on the difference between the probability of link transmission performance degradation exceeding the threshold for the first and second historical segments. This difference can be calculated by calculating the deviation between the two probabilities. If the probability of link transmission performance degradation exceeding the threshold for the first historical segment is greater than that for the second historical segment, the deviation is large, indicating that the first segment (i.e., the high-dynamic voice segment) is more likely to experience performance degradation during link transmission, and the system needs to adjust its priority accordingly. If the difference between the two probabilities is small, the link performance of the high-dynamic voice segment is not much different from that of the normal segment, and the corresponding priority adjustment will be smaller.

[0139] The significance and technical benefit of this process lies in that by comparing the probability of transmission performance degradation between the first and second categories of historical segments, the system can more accurately measure the performance differences between high-dynamic voice segments and other segments during link transmission. This allows the system to dynamically adjust the scheduling priority of these segments based on actual conditions, ensuring that high-dynamic voice segments receive priority when link conditions are poor, thereby improving audio and video transmission quality and user experience.

[0140] Furthermore, the AI-based audio and video data processing system also includes:

[0141] The second correction parameter determination module 400 is configured to calculate the reconstruction cost value corresponding to each first-category historical segment and determine the second correction parameter according to a change trend characteristic of the reconstruction cost value.

[0142] Specifically, Figure 7 FIG. 4 is a structural block diagram of a second correction parameter determination module 400 in a system provided by an embodiment of the present invention.

[0143] In a preferred embodiment of the present invention, the second correction parameter determination module 400 specifically includes:

[0144] The transmission data parsing unit 401 is configured to parse the link transmission data recorded in each first-category historical segment, extract the actual number of retransmissions and the transmission delay corresponding to each first-category historical segment, and calculate a reconstruction cost value corresponding to each first-category historical segment based on the actual number of retransmissions and the transmission delay; the actual number of retransmissions is obtained by the number of data retransmissions occurring during the link transmission process, and the transmission delay is calculated by the time difference between the data transmission time and the data reception time of each first-category historical segment; the reconstruction cost value is calculated by a weighted linear combination of the actual number of retransmissions and the transmission delay;

[0145] A change curve construction unit 402 is used to count the reconstruction cost values corresponding to all first-category historical segments and construct a reconstruction cost value change curve according to the segment time sequence;

[0146] The average slope calculation unit 403 is used to calculate the average slope of the reconstruction cost value change curve and use the average slope as the second correction parameter.

[0147] In the embodiment of the present invention, the specific working process of the transmission data parsing unit 401 is as follows:

[0148] First, the system parses the link transmission data recorded in each first-category historical segment and extracts the actual number of retransmissions and transmission delay corresponding to each segment. The actual number of retransmissions is obtained by monitoring the number of packet retransmissions during link transmission, that is, calculating the retransmission frequency of each segment. Transmission delay is obtained by calculating the time difference between the data transmission time and the reception time of each historical segment. It is usually expressed as the time required for the packet to propagate across the network.

[0149] Next, the system calculates the reconstruction cost for each first-category historical segment by performing a weighted linear combination of the actual number of retransmissions and the transmission delay. Before calculating the reconstruction cost for each first-category historical segment, the system first normalizes the number of retransmissions and the transmission delay. Because the number of retransmissions and the transmission delay have different dimensions and magnitudes, a direct weighted linear combination could result in one metric overly influential in the final result. Therefore, normalization is required to ensure that both metrics are weighted equally.

[0150] After normalization, the system weights the number of retransmissions and transmission delay according to a preset weighting factor and sums them to obtain the reconstruction cost for each segment. This weighted linear combination allows the system to comprehensively consider the impact of link stability (reflected by the number of retransmissions) and transmission delay (reflected by the transmission delay) on the reconstruction cost, thereby calculating the reconstruction difficulty and resource consumption required for each segment.

[0151] In the change curve construction unit 402, the system calculates the reconstruction cost values for all first-category historical segments and constructs a reconstruction cost change curve based on the chronological order of the segments. Specifically, the system arranges the reconstruction cost values for each historical segment in chronological order to produce a curve that reflects the change in reconstruction cost over time. This curve can demonstrate the fluctuations in reconstruction demand within each time period during the transmission process.

[0152] In average slope calculation unit 403, the system calculates the average slope of the reconstruction cost change curve and uses this average slope as the second correction parameter. By calculating the average slope of the curve, the system can determine the overall trend of the reconstruction cost over time. If the slope of the curve is positive, it means that the reconstruction cost value is gradually increasing over time, indicating that the quality of the transmission link may be deteriorating, resulting in a higher reconstruction cost. If the slope of the curve is negative, it means that the reconstruction cost is gradually decreasing, which may mean that the link quality is improving and the reconstruction cost is decreasing. This average slope value, as the second correction parameter, can reflect the fluctuation of the link during the transmission process and its impact on the reconstruction cost.

[0153] Furthermore, the AI-based audio and video data processing system also includes:

[0154] The correction module 500 is used to correct the initial scheduling priority value by combining the first correction parameter and the second correction parameter.

[0155] In this embodiment of the present invention, the purpose of combining the first and second correction parameters for comprehensive correction is to comprehensively adjust the scheduling priority of the target audio and video clips to cope with changes in the network link and the transmission requirements of different types of audio and video clips. The first correction parameter is primarily based on the real-time status of the link transmission performance, such as factors such as packet loss rate and latency. It reflects the quality of the current link and can adjust the priority of the clip in a timely manner, especially when the network fluctuates significantly during transmission. The second correction parameter is based on trends in historical data, such as the changing trend of the reconstruction cost value, which captures the fluctuation of the network link over time.

[0156] This comprehensive correction allows the system to more flexibly respond to sudden link fluctuations and long-term trends based on historical data, enabling more accurate adjustment of transmission priorities and ensuring that highly dynamic voice clips are prioritized even in unstable network environments. The combination of the first and second correction parameters provides a solution that balances real-time and historical factors, avoiding the potential bias caused by a single factor. This allows the system to more accurately determine the transmission priority of each audio and video clip, improving overall transmission efficiency and stability.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An AI-based audio and video data processing method, characterized in that: The method comprises: After determining that the target audio and video clip belongs to the high-dynamic voice characteristic type, obtaining the initial scheduling priority value generated for the target audio and video clip in the link end warning state, and obtaining the historical transmission data record of the associated device that carries the target audio and video clip transmission task; Parsing historical transmission data records, screening out a number of first-category historical segments that are consistent with the environmental characteristics of the target audio and video segments and belong to the high-dynamic voice characteristic type, and a number of second-category historical segments that do not belong to the high-dynamic voice characteristic type; Counting the probability of link transmission performance degradation exceeding a preset threshold in a number of first-category historical segments and a number of second-category historical segments, respectively, and determining a first correction parameter based on a deviation between the two; Calculating a reconstruction cost value corresponding to each first-category historical segment, and determining a second correction parameter based on a change trend characteristic of the reconstruction cost value; The steps of calculating the reconstruction cost value corresponding to each first-category historical segment and determining the second correction parameter according to the change trend characteristics of the reconstruction cost value include: Parsing the link transmission data recorded in each first-category historical segment, extracting the actual number of retransmissions and transmission delay corresponding to each first-category historical segment, and calculating the reconstruction cost value corresponding to each first-category historical segment based on the actual number of retransmissions and transmission delay; Count the reconstruction cost values corresponding to all first-category historical segments, and construct a reconstruction cost change curve according to the time sequence of the segments; Calculating the average slope of the reconstruction cost value change curve, and using the average slope as the second correction parameter; The initial scheduling priority value is corrected in combination with the first correction parameter and the second correction parameter.

2. The AI-based audio and video data processing method according to claim 1, characterized in that: The steps of respectively counting the probability of link transmission performance degradation exceeding a preset threshold in a plurality of first-category historical segments and a plurality of second-category historical segments, and determining a first correction parameter based on a deviation between the two include: Sequentially parsing each of the first-category historical segments and the second-category historical segments, extracting link transmission degradation event data recorded in each historical segment to determine the corresponding link transmission performance degradation magnitude; Count the number of segments in which the link transmission performance degradation exceeds a preset threshold in a number of first-category historical segments and a number of second-category historical segments, and calculate the corresponding occurrence probability; A first correction parameter is determined based on a deviation between the probability of occurrence that the link transmission performance degradation range in the first type of historical segments and the second type of historical segments exceeds a preset threshold standard.

3. The AI-based audio and video data processing method according to claim 2, characterized in that: The link transmission degradation event data includes packet loss rate, delay jitter amplitude and instantaneous bandwidth fluctuation amplitude. The link transmission performance degradation amplitude is calculated by weighted combination of packet loss rate, delay jitter amplitude and instantaneous bandwidth fluctuation amplitude.

4. The AI-based audio and video data processing method according to claim 1, characterized in that: The actual number of retransmissions is obtained by the number of data retransmissions that occur during the link transmission process, and the transmission delay is calculated by the time difference between the data sending time and the receiving time of each first-category historical segment; The reconstruction cost value is calculated by performing a weighted linear combination of the actual number of retransmissions and the transmission delay.

5. The AI-based audio and video data processing method according to claim 1, characterized in that: The step of correcting the initial scheduling priority value by combining the first correction parameter and the second correction parameter includes: Retrieving a preset dispatch priority value correction formula, substituting the first correction parameter and the second correction parameter into the dispatch priority value correction formula, correcting the initial dispatch priority value, and obtaining a corrected dispatch priority value; The modified scheduling priority value is applied to the link resource scheduling policy of the target video segment to dynamically adjust the scheduling priority level of the target video segment in the transmission link.

6. The AI-based audio and video data processing method according to claim 5, characterized in that: The scheduling priority value correction formula is: ; in Refers to the revised dispatch priority value; Refers to the initial scheduling priority value, Refers to the first correction parameter, that is, the deviation of the probability of link transmission performance degradation exceeding the preset threshold standard in the first and second historical segments, Refers to the adjustment coefficient corresponding to the first correction parameter, Refers to the second correction parameter, that is, the average slope of the reconstruction cost value change curve, Refers to the adjustment coefficient corresponding to the second correction parameter; In the scheduling priority value correction formula, ; in Refers to the probability of link transmission performance degradation exceeding the preset threshold standard in the first type of historical segment, Refers to the probability of link transmission performance degradation exceeding the preset threshold standard in the second type of historical fragments. It is the preset minimum protection value, used to avoid calculation anomalies when the denominator is zero or close to zero.

7. An AI-based audio and video data processing system, characterized in that: The system includes: a data acquisition module, a data screening module, a first correction parameter determination module, a second correction parameter determination module and a correction module, wherein: A data acquisition module is used to obtain, after determining that the target audio and video clip belongs to the high-dynamic voice characteristic type, an initial scheduling priority value generated for the target audio and video clip in the link end warning state, and obtain historical transmission data records of the associated device that carries the target audio and video clip transmission task; A data screening module is used to parse historical transmission data records and screen out a number of first-category historical segments that are consistent with the environmental characteristics of the target audio and video segments and belong to the high-dynamic voice characteristic type, and a number of second-category historical segments that do not belong to the high-dynamic voice characteristic type; A first correction parameter determination module is configured to calculate the probability of link transmission performance degradation exceeding a preset threshold in a plurality of first-category historical segments and a plurality of second-category historical segments, and determine a first correction parameter based on a deviation between the two. a second correction parameter determination module, configured to calculate a reconstruction cost value corresponding to each first-category historical segment, and determine a second correction parameter based on a change trend characteristic of the reconstruction cost value; The second correction parameter determination module specifically includes: a transmission data parsing unit, configured to parse the link transmission data recorded in each first-category historical segment, extract the actual number of retransmissions and the transmission delay corresponding to each first-category historical segment, and calculate a reconstruction cost value corresponding to each first-category historical segment based on the actual number of retransmissions and the transmission delay; the actual number of retransmissions being obtained by the number of data retransmissions occurring during the link transmission process, the transmission delay being calculated by the time difference between the data transmission time and the data reception time of each first-category historical segment; and the reconstruction cost being calculated by a weighted linear combination of the actual number of retransmissions and the transmission delay; A change curve construction unit is used to count the reconstruction cost values corresponding to all first-category historical segments and construct a reconstruction cost value change curve according to the time sequence of the segments; an average slope calculation unit, configured to calculate an average slope of the reconstruction cost value change curve and use the average slope as a second correction parameter; The correction module is used to correct the initial scheduling priority value by combining the first correction parameter and the second correction parameter.

8. The AI-based audio and video data processing system according to claim 7, characterized in that: The first correction parameter determination module specifically includes: a historical segment parsing unit, configured to sequentially parse each of the first and second historical segments, extract link transmission degradation event data recorded in each historical segment, and determine a corresponding link transmission performance degradation magnitude; the link transmission degradation event data includes packet loss rate, delay jitter magnitude, and instantaneous bandwidth fluctuation magnitude, and the link transmission performance degradation magnitude is calculated by performing a weighted combination of the packet loss rate, delay jitter magnitude, and instantaneous bandwidth fluctuation magnitude; an occurrence probability calculation unit, configured to count the number of segments in which the link transmission performance degradation exceeds a preset threshold value in the plurality of first-category historical segments and the plurality of second-category historical segments, and calculate the corresponding occurrence probability thereof; The deviation amplitude determining unit is used to determine a first correction parameter based on the deviation amplitude of the probability of the link transmission performance degradation amplitude exceeding a preset threshold standard in the first type of historical segments and the second type of historical segments.

Images