Sound control method based on human position detection and sound system

Abstract

The application relates to a sound control method based on human body position detection and a sound, wherein a position detection device of the sound is configured as a human body position detection hardware and is used to detect the position of a target human body in real time and obtain detection information; a processor device is provided with an associated model of any one of the above embodiments and is configured to perform similarity calculation on the detection information and position information according to the associated model, and to control a power amplifier device according to the associated logical time sequence. Through pre-calibration of the sound control logic, the logical time sequence is called according to the similarity calculation of the detection information and the position information, so that the control of the power amplifier device is closer to the ideal control effect.

Description

Technical Field

[0001] The present application relates to the field of intelligent audio technology, and in particular to an audio control method and an audio system based on human body position detection. Background Art

[0002] "Sound" is short for sound system, referring to a complete set of equipment that can reproduce and play audio signals. By receiving, amplifying, and reproducing sound signals, a sound system provides a high-quality sound experience. Sound systems typically include components such as microphones, decoders, power amplifiers, and speakers, which work together to convert audio signals into audible sound.

[0003] Currently, with the development of intelligent hardware technology, the technical form and control logic of traditional audio have also undergone a large number of improvements. For example, intelligent volume control replaces the traditional manual volume operation control method, that is, a more intelligent audio control solution. In this type of intelligent audio control solution, the use of human position information as the basis of control variables is currently relatively widely used, which can achieve improved automation of audio control and human-computer interaction. The current audio control solution based on human position detection mainly includes a human positioning device, an audio control circuit, and an audio device. The human positioning device detects the position information of the target person and converts it into a variable input for the audio control circuit in real time. The audio control circuit outputs a control signal to adjust the state of the audio device.

[0004] However, current sound control solutions based on human position detection primarily rely on real-time changes in human position for real-time control. This results in abrupt changes in control timing, poorly matching the actual needs of sound control, and only adequately addressing single control requirements. When faced with complex sound control requirements, especially those based on complex human-computer interaction, current sound control solutions based on human position detection have been ineffective. Summary of the Invention

[0005] In view of the above analysis, the embodiments of the present invention aim to provide a sound control method and sound based on human position detection, so as to solve the problem that the current sound control scheme based on human position detection has poor implementation effect in the sound control scenario based on complex human-computer interaction requirements in the existing technology implementation results.

[0006] The present invention provides a method for controlling sound based on human position detection, including:

[0007] Collecting location sample data, calibrating the sound control logic based on the location sample data, and establishing a sound control model; wherein the location sample data includes location information, the sound control logic includes a logic sequence, and the sound control model is a correlation model of the location information and the logic sequence;

[0008] Detect the position of the target body in real time and obtain detection information;

[0009] The similarity between the detection information and the position information is calculated according to the association model, and the associated logic sequence is called to control the power amplifier device.

[0010] The sound control method based on human position detection in the present embodiment collects position sample data and calibrates the sound control logic based on the position sample data to establish an sound control model. The sound control model is a correlation model between position information and logical timing. The position of the target person is detected in real time to obtain detection information. Similarity calculations are performed between the detection information and the position information based on the correlation model, and the associated logical timing is invoked to control the power amplifier device. By using pre-calibrated sound control logic and invoking the logical timing based on the similarity calculations between the detection information and the position information, the control of the power amplifier device is closer to the ideal control effect.

[0011] As one of the optional embodiments, the position sample data includes continuous position information of a first time period; the sound control logic includes continuous logic timing of a second time period;

[0012] Among them, location information and logical timing are structured data.

[0013] As one of the optional embodiments, the process of building the association model includes the following steps:

[0014] Split the location information to obtain information blocks;

[0015] Multiple logical time sequences are embedded into information blocks in a vector embedding manner to obtain an association model in the form of a vector library.

[0016] As one of the optional embodiments, a process of calculating similarity between detection information and location information according to an association model and calling an associated logic sequence to control a power amplifier device includes the following steps:

[0017] Calculate the cosine similarity between each detection information and the position information;

[0018] According to the supervised vector scheme, the cosine similarity is optimized to determine the location information that best matches the detection information.

[0019] As one of the optional embodiments, the process of calculating the cosine similarity between each detection information and the position information is as follows: ;in, A is the vector representation of the detection information, B is the vector representation of position information, yes A and B The angle between is a vector A The norm of is a vector B The norm of similarity Represents cosine similarity.

[0020] As one of the optional embodiments, the process of optimizing the cosine similarity according to the supervised vector scheme and determining the location information that best matches the detection information includes the following steps:

[0021] The CoSENT scheme is adopted to calculate the cosine similarity so that the association model focuses on improving the matching similarity score and reducing the mismatching similarity score, thereby obtaining the location information that best matches the detection information.

[0022] As one of the optional embodiments, the process of calculating similarity between the detection information and the position information according to the association model and calling the associated logic sequence to control the power amplifier device further includes the following steps:

[0023] Constraining cosine similarity via a scoring scheme.

[0024] As one of the optional embodiments, the scoring scheme is as follows: Among them, D represents information block, Q represents query, Represents a structural unit in a query, is the number of structured units contained in the information block D; k1 is an adjustable parameter, where k1=1.2-2.0, b is 0.0-1.0, len(D) is the length of the position information, and avg_len is the average length of each information block; is less than the set value, and ; similarity Represents cosine similarity.

[0025] The present application also provides an audio control device based on human position detection, including:

[0026] a model building module for collecting location sample data, calibrating the sound control logic based on the location sample data, and establishing a sound control model; wherein the location sample data includes location information, the sound control logic includes a logic sequence, and the sound control model is a correlation model of the location information and the logic sequence;

[0027] Information acquisition module, used to detect the position of the target body in real time and obtain detection information;

[0028] The control execution module is used to calculate the similarity between the detection information and the position information according to the association model, and call the associated logic sequence to control the power amplifier device.

[0029] The sound control device based on human position detection in the present embodiment collects position sample data and calibrates the sound control logic based on the position sample data to establish an sound control model. The sound control model is a correlation model between position information and logical timing. The device detects the target human's position in real time to obtain detection information. Similarity calculations are performed between the detection information and the position information based on the correlation model, and the associated logical timing is invoked to control the power amplifier device. By using pre-calibrated sound control logic and invoking the logical timing based on the similarity calculations between the detection information and the position information, the control of the power amplifier device is closer to the ideal control effect.

[0030] The present application also provides a speaker, including:

[0031] Power amplifier equipment;

[0032] A position detection device is configured to detect the position of a target human body in real time and obtain detection information;

[0033] The processor device stores the association model of any of the above embodiments and is configured to perform similarity calculation on the detection information and the position information according to the association model, and call the associated logic sequence to control the power amplifier device.

[0034] In the embodiments of the present application, the position detection device of the audio system, serving as human position detection hardware, is configured to detect the position of a target person in real time and obtain detection information. A processor device stores a correlation model from any of the aforementioned embodiments and is configured to perform a similarity calculation between the detection information and the position information based on the correlation model, invoking the associated logic sequence to control the power amplifier device. Using pre-calibrated audio control logic, the logic sequence is invoked based on the similarity calculation between the detection information and the position information, enabling control of the power amplifier device to more closely resemble ideal control results. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flow chart of a highway accident handling and tracking method based on video surveillance provided by the present invention;

[0036] Figure 2 A flow chart of a highway accident handling and tracking method based on video surveillance according to a preferred embodiment of the present invention;

[0037] Figure 3 A module structure diagram of a highway accident handling and tracking device based on video surveillance according to an embodiment of the present invention;

[0038] Figure 4 A schematic block diagram of a data control device provided by the present invention;

[0039] Figure 5A schematic diagram of a non-transitory computer-readable storage medium provided by the present invention;

[0040] Figure 6 Schematic diagram of the sound structure of a disclosed embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the embodiments of the present application will be clearly and completely described below in conjunction with the drawings of the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the described embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0042] Unless otherwise defined, the technical or scientific terms used in this application should have the usual meanings understood by people with ordinary skills in the field to which this application belongs. The "first", "second" and similar words used in this application do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0043] In order to keep the following description of the embodiments of the present application clear and concise, the present application omits detailed descriptions of some known functions and known components.

[0044] The embodiment of the present application provides an audio control method based on human body position detection.

[0045] Figure 1 FIG. 1 is a flow chart of a sound control method based on human position detection according to a disclosed embodiment. Figure 1 As shown, a sound control method based on human position detection according to a disclosed embodiment includes steps S100 to S102:

[0046] S100, collecting location sample data, and calibrating the sound control logic based on the location sample data to establish an sound control model; wherein the location sample data includes location information, the sound control logic includes a logic sequence, and the sound control model is a correlation model of the location information and the logic sequence;

[0047] S101, detecting the position of the target human body in real time and obtaining detection information;

[0048] S102: performing similarity calculation on the detection information and the position information according to the association model, and calling the associated logic sequence to control the power amplifier device.

[0049] In the disclosed embodiments, position sample data serves as training data, and its collection scenario and hardware are consistent with the audio application scenario. The hardware used to collect position sample data is a position detection device, which is used based on the audio implementation plan. By testing changes in different human positions, the required audio control logic is calibrated to achieve the control effects corresponding to the position sample data.

[0050] As one of the preferred embodiments, the position sample data includes continuous position information of a first period; the sound control logic includes continuous logic timing of a second period;

[0051] Among them, location information and logical timing are structured data.

[0052] The position information of the first time period, that is, the change in the position of the human body is collected continuously and exists in a time period, and the position information of the first time period is obtained. The audio control logic is a logical timing of continuous and existing time periods. For example, the first time period is 5s, and the second time period is 3s. In the application of audio, the position detection device detects the position of the human body for 5s continuously according to the calibration, and the detection information obtained can be converted into a continuous 3s control of the power amplifier device at the back end. The solution of the present application is different from the traditional method that changes in the control logic immediately when the position of the human body changes. It is more optimized in actual experience and avoids the bad experience caused by the abrupt change of control. At the same time, according to the calibration of the first time period and the second time period, the audio control model formed can enrich the control means according to different needs and improve the flexibility of audio control based on human position detection. Preferably, the first time period is greater than the second time period.

[0053] In the embodiments of the present application, the location information and logical timing are converted into structured data to facilitate the establishment of an association between the location information and the logical timing, forming the basis of the association model. Preferably, text data is selected as the structured data to facilitate the subsequent invocation of the relevant language model to build the association model. At the same time, in actual implementation plans, the application of language models is relatively convenient and has advantages in hardware and software costs.

[0054] Based on this, Figure 2 The flowchart of the method for building the association model of the basic embodiment is as follows: Figure 2 As shown, the process of building the association model includes steps S200 and S201:

[0055] S200, splitting the location information to obtain information blocks;

[0056] S201 , embed multiple logical time sequences into information blocks in a vector embedding manner to obtain an association model in a vector library form.

[0057] Figure 3 Build a schematic diagram for the association model, such as Figure 3 As shown, to address the discrepancy between the first and second time periods, the location information is split into information blocks, into which the logical timing is embedded using vector embedding, forming an association model in the form of a vector library. This disclosed embodiment innovatively introduces multiple parallel vector embedding methods to mitigate the impact of single logical timing embedding on the information blocks (position-to-vector encoding). Furthermore, in the application of the association model, a scheme such as cosine similarity is used to compare detection information with location information. This comparison determines the location information and invokes the corresponding logical timing.

[0058] Preferably, reference Figure 3 ,After the detection information is collected, it is also vectorized in the same way as the ,location information. After it is converted into structured data, the ,association model is imported to perform similarity calculation with the ,location information, and the matching logical sequence is called based on the ,most similar location information. Figure 4 FIG. 1 is a flow chart of a sound control method based on human position detection according to a preferred embodiment. Figure 4 As shown, in step S102, similarity calculation is performed on the detection information and the position information according to the association model, and the process of calling the associated logic sequence to control the power amplifier device includes steps S1020 and S1021:

[0059] S1020, calculating the cosine similarity between each detection information and the position information;

[0060] S1021, optimizing the cosine similarity according to the supervised vector scheme to determine the location information that best matches the detection information.

[0061] The process of calculating the cosine similarity between each detection information and the position information is as follows: ;in, A Vector representation of detection information, B is the vector representation of position information, yes A and B The angle between and is a vector A The norm of is a vector B The norm of similarity Represents cosine similarity.

[0062] In the association model, there are multiple position information. The detection information is subjected to cosine similarity calculation with each position information. The most similar position information is determined based on the cosine similarity, and the logic sequence corresponding to the most similar position information is called to perform audio control.

[0063] Preferably, the process of optimizing the cosine similarity according to the supervised vector scheme in step S1021 to determine the location information that best matches the detection information includes the following steps:

[0064] The CoSENT scheme is adopted to calculate the cosine similarity so that the association model focuses on improving the matching similarity score and reducing the mismatching similarity score, thereby obtaining the location information that best matches the detection information.

[0065] Using supervised vector solutions such as CoSENT, the accuracy of similarity tasks is improved by optimizing cosine similarity. Actual tests of the audio control solution show that CoSENT's training process is more consistent with the prediction task because it directly optimizes cosine similarity rather than indirectly affecting similarity through classification tasks. Since the task of detecting information and location information in this application is more direct, the CoSENT solution is more effective. The CoSENT solution is as follows: in, 、 is the set of all positive and negative sample pairs, that is, the set of correctly matched vector samples and the set of mismatched vector samples. It is a hyperparameter, usually set to 20. The purpose of the loss function is to make the cosine similarity of any match in a batch greater than the cosine similarity of the unmatched ones, and through the calculation of cosine similarity, allow the association model to focus on improving the similarity score of the match and ensure that the similarity score of the unmatched ones is reduced.

[0066] At the same time, the CoSENT solution was selected to further improve the convenience of implementing the processor device. In actual deployment, the large language model was used to build and use the association model, and the solution achieved excellent application results and low costs.

[0067] Preferably, in the embodiment of the present disclosure, Figure 4 As shown, the process of calculating the similarity between the detection information and the position information according to the association model in step S102 and calling the associated logic sequence to control the power amplifier device also includes step S1022:

[0068] S1022, constraining cosine similarity via a scoring scheme.

[0069] The grading scheme is as follows: Among them, D represents information block, Q represents query, Represents a structural unit in a query, is the number of structured units contained in the information block D; k1 is an adjustable parameter, where k1=1.2-2.0, b is 0.0-1.0, len(D) is the length of the position information, and avg_len is the average length of each information block; from IDF We can see from the formula that The larger the denominator, the smaller the numerator, that is, the smaller the IDF The smaller the value, the smaller the value. N represents the number of similar information blocks in the information block D, that is, how many information blocks are there in total to calculate the cosine similarity with the detection information, and 0.5 is the smoothing factor, which is a hyperparameter used to prevent q i The denominator is 0.

[0070] in, is less than the set value, and ; similarity Represents cosine similarity.

[0071] After the logic sequence of the call is determined, the relevant logic sequence can be used to control the power amplifier equipment, such as adjusting the parameters of the audio, switching, etc.

[0072] The sound control method based on human position detection in the disclosed embodiments collects position sample data and calibrates the sound control logic based on the position sample data to establish an sound control model. The sound control model is a correlation model between position information and logical timing. The position of the target person is detected in real time to obtain detection information. Similarity calculations are performed between the detection information and the position information based on the correlation model, and the associated logical timing is invoked to control the power amplifier device. Using pre-calibrated sound control logic, the logical timing is invoked based on the similarity calculations between the detection information and the position information, resulting in a more ideal control effect for the power amplifier device.

[0073] Any embodiment of the present disclosure further provides an audio control device based on human body position detection.

[0074] Figure 5 FIG. 1 is a block diagram of a sound control device based on human position detection according to a disclosed embodiment. Figure 5 Shown, including:

[0075] The model building module 100 is used to collect location sample data, calibrate the sound control logic based on the location sample data, and establish an sound control model; wherein the location sample data includes location information, the sound control logic includes a logical sequence, and the sound control model is a correlation model of the location information and the logical sequence;

[0076] The information acquisition module 101 is used to detect the position of the target body in real time and obtain detection information;

[0077] The control execution module 102 is used to perform similarity calculation on the detection information and the position information according to the association model, and call the associated logic sequence to control the power amplifier device.

[0078] The sound control device based on human position detection in the disclosed embodiments collects position sample data and calibrates the sound control logic based on the position sample data to establish an sound control model. The sound control model is a correlation model between position information and logical timing. The device detects the target human's position in real time to obtain detection information. Similarity calculations are performed between the detection information and the position information based on the correlation model, and the associated logical timing is invoked to control the power amplifier. Using pre-calibrated sound control logic, the logic timing is invoked based on the similarity calculations between the detection information and the position information, resulting in a more realistic control effect for the power amplifier.

[0079] Any embodiment of the present disclosure further provides a speaker.

[0080] Figure 6 FIG. 1 is a schematic diagram of an acoustic structure of a disclosed embodiment, such as Figure 6 Shown, including:

[0081] Power amplifier device 200;

[0082] The position detection device 201 is configured to detect the position of the target human body in real time and obtain detection information;

[0083] The position detection device 201 may use a distance measuring device, such as an infrared rangefinder, and the obtained detection information is structured and converted in the processor device 202 .

[0084] The processor device 202 stores the association model of any of the above embodiments and is configured to perform similarity calculation on the detection information and the position information according to the association model, and call the associated logic sequence to control the power amplifier device 200.

[0085] In the embodiments of the present application, the position detection device of the audio system, serving as human position detection hardware, is configured to detect the position of a target person in real time and obtain detection information. A processor device stores a correlation model from any of the aforementioned embodiments and is configured to perform a similarity calculation between the detection information and the position information based on the correlation model, invoking the associated logic sequence to control the power amplifier device. Using pre-calibrated audio control logic, the logic sequence is invoked based on the similarity calculation between the detection information and the position information, enabling control of the power amplifier device to more closely resemble ideal control results.

[0086] Regarding this application, the following points need to be explained:

[0087] (1) The drawings of the embodiments of this application only relate to the structures related to the embodiments of this application. Other structures can refer to the general design.

[0088] (2) For the sake of clarity, the thickness and size of layers or structures in the drawings used to describe the embodiments of the present invention are exaggerated. It will be understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element can be "directly on" or "under" the other element, or intervening elements may be present.

[0089] (3) Unless there is a conflict, the embodiments of this application and the features therein may be combined to form new embodiments. The above are only specific implementation methods of this application, but the scope of protection of this application is not limited thereto. The scope of protection of this application shall be based on the scope of protection of the claims.

[0090] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0091] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A sound control method based on human body position detection, characterized in that: Including steps: Collecting location sample data, calibrating the sound control logic based on the location sample data, and establishing a sound control model; wherein the location sample data includes location information, the sound control logic includes a logic sequence, and the sound control model is a correlation model of the location information and the logic sequence; Detect the position of the target body in real time and obtain detection information; performing similarity calculation on the detection information and the location information according to the association model, and calling the associated logic sequence to control the power amplifier device; The position sample data includes continuous position information of a first period; the sound control logic includes continuous logic sequence of a second period; wherein the position information and the logic sequence are structured data; The process of building the association model includes the following steps: Split the location information to obtain information blocks; Embed multiple logical time sequences into information blocks in a vector embedding manner to obtain an association model in the form of a vector library; After the detection information is collected, it is vectorized in the same way as the location information. After being converted into structured data, it is imported into the association model and similarity is calculated with the location information. The matching logic sequence is called based on the most similar location information. The process of performing similarity calculation on the detection information and the position information according to the association model and calling the associated logic sequence to control the power amplifier device includes the steps of: Calculate the cosine similarity between each detection information and the position information; The CoSENT scheme is adopted, and through the calculation of cosine similarity, the association model is focused on improving the similarity score of matches and reducing the similarity score of mismatches, thereby obtaining the position information that best matches the detection information.

2. The sound control method based on human body position detection according to claim 1, characterized in that: The process of calculating the cosine similarity between each detection information and the position information is as follows: ;in, A is the vector representation of the detection information, B is the vector representation of position information, yes A and B The angle between is a vector A The norm of is a vector B The norm of similarity Represents cosine similarity.

3. The sound control method based on human body position detection according to claim 1, characterized in that: The process of calculating similarity between the detection information and the location information according to the association model and calling the associated logic sequence to control the power amplifier device further includes the steps of: constraining the cosine similarity via a scoring scheme; The grading scheme is as follows: ; Where D stands for information block and Q stands for query. Represents a structural unit in a query, is the number of structured units contained in the information block D; k1 is an adjustable parameter, where k1=1.2-2.0, b is 0.0-1.0, len(D) is the length of the position information, and avg_len is the average length of each information block; Where N represents the number of similar information blocks in the information block D, that is, the total number of information blocks and the detection information to calculate the cosine similarity; Where, is less than the set value, and ; similarity Represents cosine similarity.

4. A sound control device based on human body position detection, characterized in that: include: a model building module, configured to collect location sample data, calibrate the sound control logic based on the location sample data, and establish a sound control model; wherein the location sample data includes location information, the sound control logic includes a logic sequence, and the sound control model is a correlation model of the location information and the logic sequence; Information acquisition module, used to detect the position of the target body in real time and obtain detection information; a control execution module, configured to perform similarity calculation on the detection information and the position information according to the association model, and call the associated logic sequence to control the power amplifier device; The position sample data includes continuous position information of a first period; the sound control logic includes continuous logic timing of a second period; Among them, location information and logical timing are structured data; The process of building the association model includes the following steps: Split the location information to obtain information blocks; Embed multiple logical time sequences into information blocks in a vector embedding manner to obtain an association model in the form of a vector library; After the detection information is collected, it is vectorized in the same way as the location information. After being converted into structured data, it is imported into the association model and similarity is calculated with the location information. The matching logic sequence is called based on the most similar location information. The process of performing similarity calculation on the detection information and the position information according to the association model and calling the associated logic sequence to control the power amplifier device includes the steps of: Calculate the cosine similarity between each detection information and the position information; The CoSENT scheme is adopted, and through the calculation of cosine similarity, the association model is focused on improving the similarity score of matches and reducing the similarity score of mismatches, thereby obtaining the position information that best matches the detection information.

5. A sound system, characterized in that: include: Power amplifier equipment; A position detection device is configured to detect the position of a target human body in real time and obtain detection information; A processor device stores the association model according to any one of claims 1 to 3, and is configured to perform similarity calculation on the detection information and the position information according to the association model, and call the associated logic sequence to control the power amplifier device.

Images