Intelligent audio-driven fragrance control system, control method and vehicle-mounted system

Abstract

The invention discloses an intelligent audio-driven fragrance control system, a control method and a vehicle-mounted system. The method comprises the following steps: receiving a real-time audio signal; decomposing the real-time audio into a time sequence set of a plurality of music elements by adopting a neural network model; determining the type and signal characteristics (volume, frequency, rhythm and the like) of each music element; according to a preset matching rule, determining a fragrance base tone matched with the type of the music element; a fragrance release control instruction (release concentration, release frequency, release power and the like) is generated according to the signal characteristics of the music elements; a target fragrance container is determined according to the matched fragrance basic tone, and the valve opening degree, the valve switching frequency, the valve switching speed and the like of a fragrance release opening corresponding to the target fragrance container are controlled according to the generated fragrance release control instruction. Dynamic linkage of decomposition of multiple music elements and fragrance is achieved through the AI audio analysis technology, and the concentration and combination of the fragrance can be accurately controlled according to the rhythm, frequency and other characteristics of the music.

Description

Technical Field

[0001] The present invention relates to the field of intelligent fragrance driving, and particularly to an intelligent audio-driven fragrance control system, a control method, and a vehicle-mounted system. Background Art

[0002] With the continuous growth of consumers' sensory needs, fragrance systems are widely used in vehicle-mounted, home, and commercial scenarios to enhance the environmental atmosphere and users' immersive experience.

[0003] Traditional fragrance technologies are used to manually adjust the fragrance concentration to provide basic functions, and some also automatically spray fragrance by setting fixed time intervals, which are suitable for simple environmental fragrance adjustment. Although these two methods are simple in implementation, they cannot respond to real-time changes, and the user experience is relatively single.

[0004] To further meet users' personalized needs for fragrance, some fragrances are built-in with multiple scents, and users can switch the release of different scents by selecting preset modes. For example, the soothing mode corresponds to the release of lavender fragrance, and the fresh mode corresponds to the release of jasmine fragrance, etc. However, this preset mode is usually statically set and is difficult to adapt to users' real-time needs.

[0005] To achieve adaptability to real-time needs, some intelligent fragrance systems adjust the type or concentration of fragrance through environmental sensors such as temperature sensors, humidity sensors, and air quality sensors. For example, vehicle-mounted fragrances that switch different scents according to the type of music. For example, Chinese Patent Application No. CN117068080A discloses using a deep neural network to classify music. For example, when playing classical music, it switches to sandalwood scent, and when playing DJ music, it switches to lemon scent.

[0006] The disclosure of the above background art content is only for assisting in understanding the concept and technical solution of the present application, and it does not necessarily belong to the prior art of the present application, nor does it necessarily provide technical guidance; in the case where there is no clear evidence indicating that the above content was publicly available before the filing date of the present application, the above background art should not be used to evaluate the novelty and inventiveness of the present application. Summary of the Invention

[0007] The object of the present invention is to provide an intelligent audio-driven intelligent fragrance control system based on a neural network model, which realizes the dynamic response adjustment of fragrance with the music elements decomposed from the audio, matches the fragrance keynote with fine audio granularity, and provides users with a mixed fragrance experience that is adapted to the audio and rich in elements.

[0008] To achieve the above object, the technical solution adopted by the present invention is as follows: An intelligent audio-driven fragrance control system includes a fragrance device, an audio input module, and an AI audio-driven module. Among them, the fragrance device is configured with a controller and multiple fragrance containers having fragrance release ports. Each fragrance container is configured to hold perfume with a preset fragrance base note, and each fragrance release port is provided with a control valve controlled by the controller; The audio input module is configured to receive real-time audio signals and transmit them to the AI audio-driven module; The AI audio-driven module is configured to: Decompose the real-time audio into a time series set of complex music elements using a neural network model; Determine the type and signal characteristics of each music element, where the signal characteristics include one or more of volume, frequency, and rhythm information; According to a preset matching rule, determine the fragrance base note matched by the type of the music element; and generate a fragrance release control instruction according to the signal characteristics of the music element, where the instruction includes one or more of release concentration, release frequency, and release power; The controller is configured to determine a target fragrance container according to the matched fragrance base note, and control one or more of the valve opening degree, valve switching frequency, and valve switching speed of the fragrance release port corresponding to the target fragrance container according to the generated fragrance release control instruction.

[0009] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the system further includes a user interaction module, which is configured to manually set a real-time period; The AI audio-driven module counts the types and distribution ratios of music elements within a period with the current moment as the end point of the real-time period, and according to the fragrance base notes matched by the counted types of music elements; and generates corresponding fragrance release control instructions for the various fragrance base notes matched according to the counted distribution ratios, where the lower the distribution ratio, the lower the release concentration, release frequency, and / or release power of the corresponding fragrance base note; The controller controls the valve opening degree, valve switching frequency, and / or valve switching speed of the corresponding fragrance container according to the fragrance release control instructions corresponding to the various fragrance base notes.

[0010] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the user interaction module is further configured to set a cycle duration. Every time a cycle duration elapses, the AI audio-driven module generates corresponding fragrance release control instructions for the various fragrance base notes matched according to the real-time period, and the cycle duration is less than or equal to the real-time period.

[0011] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the user interaction module further includes a music element volume adjustment unit, which is configured to manually adjust the volume values of one or more music elements; According to the adjustment operation of the music element volume adjustment unit, the AI audio driving module updates the release concentration parameter in the fragrance release control instruction.

[0012] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the user interaction module further includes a fragrance adjustment unit, which is configured to manually adjust the current fragrance base type, release concentration, release frequency, and / or release power; According to the adjustment operation of the fragrance adjustment unit, the controller adjusts the valve opening degree, valve switching frequency, and / or valve switching speed of the corresponding fragrance container.

[0013] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, according to the adjustment operation of the fragrance adjustment unit, the AI audio driving module analyzes the user's preferences, and optimizes the matching rule between the type of music element and the fragrance base, and / or the relationship between the signal characteristics of the music element and the fragrance release control instruction according to the analysis result.

[0014] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the system further includes an environmental sensor, which is configured to detect one or more information of temperature and humidity; The AI audio driving module is configured to generate the fragrance release control instruction in combination with the detection result of the environmental sensor.

[0015] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the fragrance container is configured with a driving unit for driving the change of the orientation of the fragrance release port, and the driving unit is connected to the controller; The AI audio driving module is configured to generate a control instruction for the release port orientation according to the signal characteristics of the music element, and the controller controls the driving unit to act according to the control instruction for the release port orientation.

[0016] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the user interaction module is an in-vehicle intelligent operation device or an application client on a smart mobile terminal.

[0017] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the AI audio driving module decomposes the real-time audio into a time sequence set of a plurality of music elements in the following manner: Collect the separated waveforms of multiple types of musical instruments and the full waveforms of mixed audio, and manually label classification tags for the separated waveforms and full waveforms respectively as the learning sample sets for model training and validation; Input the learning sample sets into an initial model based on a convolutional neural network and a long short-term memory neural network, learn to extract the time-frequency features of musical instruments, and implement multi-objective learning of the initial model through the time-frequency masking method to obtain an encoder-decoder model capable of predicting separated waveforms; Use the encoder-decoder model to predict the separated waveforms corresponding to multiple single sound sources arranged in sequence in the real-time audio; Identify the sound source attributes matched by the separated waveforms, where the sound source attributes include one or more of loudness, frequency, and pitch; Based on the separated waveforms of each single sound source obtained by parsing and combined with the sound source attributes matched thereto, obtain each decomposed music element.

[0018] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the initial model is configured with multiple CNN layers and multiple LSTM layers, and the CNN layers are used to extract the features of each section in the input real-time audio signal, generate the CNN features of each section, and perform spatial modeling on each section in the input signal based on the CNN features; Use the LSTM layers to process the CNN features to generate the LSTM features of each section, and perform temporal modeling on the CNN features and each section based on the LSTM features; The initial model is jointly trained based on the information after completing spatial modeling and temporal modeling to extract the time-frequency features of musical instruments to generate the encoder-decoder model.

[0019] According to another aspect of the present invention, there is provided an intelligent audio-driven fragrance control method based on a neural network model, including the following steps: Receive a real-time audio signal; Use a neural network model to decompose the real-time audio into a sequential set of complex musical elements; Determine the types and signal features of each musical element, where the signal features include one or more of volume, frequency, and rhythm information; According to a preset matching rule, determine the fragrance base tone matched by the type of the musical element; and generate a fragrance release control instruction according to the signal features of the musical element, where the instruction includes one or more of release concentration, release frequency, and release power; Determine the target fragrance container according to the matched fragrance base tone, and control one or more of the valve opening degree, valve switching frequency, and valve switching speed of the fragrance release port corresponding to the target fragrance container according to the generated fragrance release control instruction.

[0020] According to another aspect of the present invention, a vehicle-mounted system is provided, including a vehicle, a vehicle-mounted audio set inside the vehicle, and the intelligent audio-driven fragrance control system as described above. The vehicle-mounted audio is configured to play the real-time audio signal received by the audio input module of the intelligent audio-driven fragrance control system.

[0021] Further, based on any one of the foregoing technical solutions or a combination of multiple technical solutions, the AI audio driving module of the intelligent audio-driven fragrance control system is further configured to: Obtain the internal space information of the vehicle and determine the location information of the fragrance device in the internal space of the vehicle; Generate a fragrance release control instruction according to the location information of the fragrance device.

[0022] The beneficial effects brought by the technical solutions provided by the present invention are as follows: a. Through AI audio analysis and neural network models, multiple signal elements of music are separated in real time, and an accurate mapping relationship is established between them and the fragrance notes, enabling the fragrance release to dynamically respond to the characteristics of music such as volume, frequency, and rhythm, and realizing dynamic fragrance adjustment along with the flow of music; b. By deeply integrating the music content with the fragrance notes, using the fine granularity of many music elements separated from the music to match the fragrance notes, and combining the fragrance concentration, release frequency, and release power, users can feel the unique fragrance of each piece of music, thereby enhancing the multi-sensory linkage experience of the vehicle-mounted entertainment system and creating a more emotionally appealing driving atmosphere; c. Through the user interface (such as the vehicle-mounted screen or the mobile APP), users are allowed to freely adjust the fragrance intensity and ratio to meet different scenarios and emotional needs, enhancing the personalization of the system and the user's sense of participation; d. Using AI-driven signal processing technology and advanced fragrance hardware, precise proportion adjustment, concentration control, and synchronous release of multiple fragrances are achieved to ensure that the fragrance performance is natural, rich, and in line with the music changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0024] Figure 1 The schematic diagram of intelligent audio-driven fragrance control provided for an exemplary embodiment of the present invention; Figure 2Schematic block diagram of the intelligent audio-driven fragrance control system provided for an exemplary embodiment of the present invention; Figure 3 Schematic diagram of the control flow from audio processing to fragrance release provided for an exemplary embodiment of the present invention; Figure 4 Schematic diagram of the functional architecture of the user interaction module in the audio-driven fragrance control system provided for an exemplary embodiment of the present invention; Figure 5 Schematic diagram of the operation interface for adjusting the fragrance type by using the in-vehicle intelligent operation screen or APP interface to adjust parameters provided for an exemplary embodiment of the present invention; Figure 6 Schematic block diagram of the intelligent audio-driven fragrance control method provided for an exemplary embodiment of the present invention. Detailed implementation manners

[0025] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or equipment comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or equipment.

[0027] Although partial intelligence is achieved by the linkage between the sensor and the intelligent fragrance, its response logic is to link the fragrance with the environmental detection parameters, and it is impossible to link other sensory contents, which limits the depth and richness of the immersive experience.

[0028] In order to enhance the richness of the multi-sensory experience, attempts are made to provide an application technology that takes both audio and fragrance into account, such as a switch that controls the release of fragrance based on the start and stop of music playback, or simply classifying music tracks and classifying fragrance smells accordingly. However, the linkage between this type of fragrance release and music is merely a blunt and rough classification association, and the lack of depth in the linkage limits the depth and richness of the multi-sensory immersive experience. The fixed combination of music classification and fragrance classification cannot be personalized according to personal preferences, making it difficult to achieve a customized and unique experience.

[0029] In an embodiment of the present invention, an intelligent audio-driven fragrance control system based on a neural network model is proposed, which is particularly suitable for vehicle-mounted scenes. The principle of controlling fragrance is as follows: Figure 1 As shown: receiving the input signal of the in-car music, separating the input signal in real time through a neural network, then determining the correlation between the separated music elements and the fragrance tone, and generating a fragrance release instruction based on the signal of the separated music elements (such as volume, audio, rhythm, etc.); finally, the fragrance release hardware executes the release instruction to dynamically release the corresponding fragrance, and finally realizes the deep linkage between music and fragrance, providing users with a personalized, multi-sensory immersive experience.

[0030] like Figure 2 As shown, an embodiment of the present invention provides an intelligent audio-driven fragrance control system based on a neural network model, comprising a fragrance device, an audio input module and an AI audio driving module, wherein the fragrance device is configured with a controller and a plurality of fragrance containers with fragrance release ports, each fragrance container is configured to load perfume with a preset fragrance tone, and each fragrance release port is provided with a control valve controlled by the controller; The audio input module is configured to receive real-time audio signals and support multiple input sources, such as USB, Bluetooth, streaming media, etc.; and transmit them to the AI ​​audio driver module; The AI ​​audio driver module is configured as follows: Decomposing the real-time audio into a time-series set of a plurality of music elements using a neural network model; Determine the type and signal characteristics of each music element, wherein the signal characteristics include one or more of volume, frequency, and rhythm information; wherein the type of music element may be, for example, vocals, piano, bass, drums, string music, etc., and analyze the signal characteristics of each music element, such as volume, frequency, rhythm, etc.; According to the preset matching rules, determine the fragrance base tone that matches the type of the music element; and generate a fragrance release control instruction according to the signal characteristics of the music element, where the instruction includes one or more of release concentration, release frequency, and release power; specifically, the preset matching rules are actually the mapping relationship between music elements and fragrance base tones. For example, elements such as drumbeats / guitars are associated with woody scents, elements such as strings / pianos are associated with floral scents, elements of bass are associated with oriental musk, elements of electronic synthesizers are associated with citrus scents, and elements of vocals are associated with sandalwood, etc.

[0031] The controller is configured to determine the target fragrance container according to the matched fragrance base tone, and control one or more of the valve opening degree, valve switching frequency, and valve switching speed of the fragrance release port corresponding to the target fragrance container according to the generated fragrance release control instruction; specifically, for example, if it is parsed that the volume of music element A is the largest, then control the valve opening degree of the fragrance container corresponding to the fragrance associated with music element A to be larger; if it is parsed that the frequency of music element B is the largest, then control the valve switching frequency of the fragrance container corresponding to the fragrance associated with music element B to be higher; if it is parsed that the rhythm of music element C is the strongest, then control the valve switching speed of the fragrance container corresponding to the fragrance associated with music element C to be faster.

[0032] The core inventive points of the present invention are as follows: First, match the fragrance base tone with music elements as the granularity, replacing the matching of fragrances with the music style of the whole song in the prior art, making the driving granularity of audio for fragrance more delicate. Many music elements can be separated from a piece of music. As the prelude, chorus, and ending parts of the whole piece of music are played, different fragrance driving schemes are adjusted and changed in real time to adapt to the music elements, and a fragrance experience that fluctuates with the music is felt. Taking the prior art CN117068080A as an example for comparison, for a piece of music identified as classical music, the prior art will match the sandalwood fragrance for its playing time, but the technical solution of the embodiment of the present invention uses music elements as music granules, making the dynamics of audio driving fragrance richer; Second, not only match the fragrance base tone with music elements, but also combine the control of fragrance concentration, release frequency, and release power to adapt to the volume, audio, and rhythm of the music. Users can feel a unique fragrance that fits the music appropriately. Similarly, taking CN117068080A as an example, the prior art does not disclose the technology of using the volume, audio, and rhythm of music as further parameters for controlling fragrance release. Combining the above two aspects, this embodiment provides an extreme experience of music-driven fragrance.

[0033] Figure 3 Show the process from audio input to separation, mapping, and generation of fragrance release control signals, which also supports AI fragrance type adjustment, and supports adjusting music elements or other parameters through the user interaction module, such as the client APP interface on in-vehicle intelligent operation devices or intelligent mobile terminals, to adjust the fragrance type, such as Figure 5As shown, a final fragrance release configuration plan is generated according to the adjustment result.

[0034] As Figure 4 shown, the system further includes a user interaction module, which may have the following functions: Function 1: Manually set the real-time period T1. The AI audio driving module counts the types and distribution ratios of music elements within the period with the current moment as the end point of the real-time period, and matches the fragrance base tones according to the types of the counted music elements; and according to the counted distribution ratios, generates corresponding fragrance release control instructions for the various matched fragrance base tones. Among them, the lower the distribution ratio, the lower the release concentration, release frequency, and / or release power of the corresponding fragrance base tone; The controller controls the valve opening degree, valve switching frequency, and / or valve switching speed of the corresponding fragrance container according to the fragrance release control instructions corresponding to the various fragrance base tones.

[0035] Function 2: Set the cycle duration T2. Every time a cycle duration elapses, the AI audio driving module generates corresponding fragrance release control instructions for the various matched fragrance base tones according to the real-time period. T2 is less than or equal to T1. In one embodiment, T2 = T1 or T2 = T1 / 2.

[0036] Function 3: Equipped with a music element volume adjustment unit for manually adjusting the volume values of one or more music elements; for example, the music element volume adjustment unit adjusts and increases the volume of the piano type music element, and the element of the piano is associated with floral fragrance. Therefore, the AI audio driving module updates the release concentration parameter of the floral fragrance in the fragrance release control instructions, that is, increases the valve opening degree of the fragrance container corresponding to the floral fragrance.

[0037] Function 4: Equipped with a fragrance adjustment unit for manually adjusting the current types of fragrance base tones, release concentration, release frequency, and / or release power; according to the adjustment operation of the fragrance adjustment unit, the controller adjusts the valve opening degree, valve switching frequency, and / or valve switching speed of the corresponding fragrance container.

[0038] Function 5: On the basis of Function 4, according to the adjustment operation of the fragrance adjustment unit, the AI audio driving module analyzes the user's preferences, and optimizes the matching rules between the types of music elements and the fragrance base tones and / or the relationship between the signal characteristics of the music elements and the fragrance release control instructions according to the analysis results.

[0039] Function 6: When applied in the vehicle field, it can also determine the location information of the fragrance device in the interior space of the vehicle according to the interior space information of the vehicle. For example, it is determined through the APP interface that the fragrance device is set at the front windshield or in the middle position between the driver's seat and the passenger seat, and then according to the location information of the fragrance device, fragrance release control instructions are generated.

[0040] In one embodiment of the present invention, the system further includes an environmental sensor configured to detect one or more types of information such as temperature and humidity; the AI audio driving module is configured to generate the fragrance release control instruction in combination with the detection result of the environmental sensor. Specifically, for example, taking 25 °C as the reference temperature, the temperature is divided into multiple levels. The higher the temperature is above the reference temperature, the valve opening is controlled to decrease, the valve switching frequency and / or the valve switching speed are reduced; the same applies to humidity. When the humidity exceeds the reference, the diffusion range of the fragrance release needs to be reduced.

[0041] In a specific embodiment, the fragrance container is configured with a driving unit for driving the change in the orientation of the fragrance release port, and the driving unit is connected to the controller; The AI audio driving module is configured to generate a control instruction for the orientation of the release port according to the signal characteristics of the music elements, and the controller controls the driving unit to act according to the control instruction for the orientation of the release port. For example, under the drive of the driving unit, the release port rotates 360° along the horizontal plane, and the controller can control its rotation speed, the period of intermittent pause of rotation, etc.

[0042] The following will describe a specific embodiment in which the AI audio driving module decomposes real-time audio into a time sequence set of complex music elements: Collect the separated waveforms of multiple types of musical instruments and the full waveform of the mixed audio, and manually label classification labels for the separated waveforms and the full waveform respectively as the learning sample set for model training and verification; Input the learning sample set into an initial model based on a convolutional neural network and a long short-term memory neural network, learn to extract the time-frequency features of the musical instruments, and implement multi-object learning of the initial model through the time-frequency masking method to obtain an encoder-decoder model capable of predicting the separated waveforms; Use the encoder-decoder model to predict the separated waveforms corresponding to multiple single sound sources arranged in time sequence in the real-time audio; Identify the sound source attributes matched by the separated waveforms, and the sound source attributes include one or more of loudness, frequency, and pitch; According to the separated waveforms of each single sound source obtained by parsing, and in combination with the sound source attributes matched with it, obtain each decomposed music element.

[0043] Among them, the initial model is configured with multiple CNN layers and multiple LSTM layers. The CNN layers are used to extract the features of each section in the input real-time audio signal, generate the CNN features of each section, and perform spatial modeling on each section in the input signal based on the CNN features; Process the CNN features using the LSTM layer to generate LSTM features for each section, and perform temporal modeling on the CNN features and each section based on the LSTM features; The initial model is jointly trained based on the information that has completed spatial modeling and temporal modeling to extract the time-frequency features of the musical instrument, so as to generate the encoder-decoder model.

[0044] The technical solution of the AI audio driving module using a neural network model to decompose real-time audio into a complex number of musical elements is described in detail in the Chinese patent application with the publication number CN117295004A, which is incorporated into this embodiment by reference in its entirety here and will not be elaborated further.

[0045] An embodiment of the present invention also provides an intelligent audio-driven fragrance control method based on a neural network model, as Figure 6 shown, the control method includes the following steps: Receive real-time audio signals; Use a neural network model to decompose the real-time audio into a sequential set of a complex number of musical elements; Determine the type and signal characteristics of each musical element, and the signal characteristics include one or more of volume, frequency, and rhythm information; According to the preset matching rules, determine the fragrance base tone matched by the type of the musical element; and generate a fragrance release control instruction according to the signal characteristics of the musical element, and the instruction includes one or more of release concentration, release frequency, and release power; Determine the target fragrance container according to the matched fragrance base tone, and control one or more of the valve opening degree, valve switching frequency, and valve switching speed of the fragrance release port corresponding to the target fragrance container according to the generated fragrance release control instruction.

[0046] It should be noted that the intelligent audio-driven fragrance control method based on a neural network model provided in the embodiment of the present invention and the intelligent audio-driven fragrance control system based on a neural network model provided in the above embodiment belong to the same inventive concept. Here, by way of full-text citation, all the content of the embodiment of the intelligent audio-driven fragrance control system based on a neural network model is incorporated into the embodiment of the intelligent audio-driven fragrance control method based on a neural network model.

[0047] An embodiment of the present invention also provides a vehicle-mounted system, including a vehicle, a vehicle-mounted audio installed inside the vehicle, and the audio-driven fragrance control system as described above, and the vehicle-mounted audio is configured to play the real-time audio signal received by the audio input module of the audio-driven fragrance control system.

[0048] Furthermore, the AI audio driving module of the audio-driven fragrance control system is further configured to: Obtain the internal space information of the vehicle and determine the location information of the fragrance device in the internal space of the vehicle; Generate a fragrance release control instruction according to the location information of the fragrance device.

[0049] The present invention realizes the dynamic linkage between music and fragrance through AI audio analysis technology, and can accurately control the concentration and combination of fragrance according to the characteristics of music such as rhythm and frequency.

[0050] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including said element.

[0051] The above are only specific embodiments of the present application. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. An intelligent audio-driven fragrance control system, characterized in that, It includes an aroma device, an audio input module, and an AI audio driving module. Among them, the aroma device is configured with a controller and multiple aroma containers with aroma release ports. Each aroma container is configured to hold perfume with a preset fragrance base note, and each aroma release port is provided with a control valve controlled by the controller; The audio input module is configured to receive a real-time audio signal and transmit it to the AI audio driving module; The AI audio driving module is configured to: Decompose the real-time audio into a time series set of complex music elements using a neural network model; Determine the type and signal characteristics of each music element, where the signal characteristics include one or more of volume, frequency, and rhythm information; According to a preset matching rule, determine the fragrance base note matched by the type of the music element; and generate an aroma release control instruction according to the signal characteristics of the music element, where the instruction includes one or more of release concentration, release frequency, and release power; The controller is configured to determine a target aroma container according to the matched fragrance base note and control one or more of the valve opening degree, valve switching frequency, and valve switching speed of the aroma release port corresponding to the target aroma container according to the generated aroma release control instruction.

2. The intelligent audio-driven fragrance control system according to claim 1, characterized in that, The system further includes a user interaction module, which is configured to manually set a real-time period; The AI audio driving module counts the types and distribution ratios of music elements within a time period with the current moment as the end point of the real-time period, and according to the fragrance base notes matched by the counted types of music elements; and generates corresponding aroma release control instructions for the various fragrance base notes matched according to the counted distribution ratios, where the lower the distribution ratio, the lower the release concentration, release frequency, and / or release power of the corresponding fragrance base note; The controller controls the valve opening degree, valve switching frequency, and / or valve switching speed of the corresponding aroma container according to the aroma release control instructions corresponding to the various fragrance base notes.

3. The intelligent audio-driven fragrance control system according to claim 2, wherein The user interaction module is further configured to set a cycle duration. Every time a cycle duration elapses, the AI audio driving module generates corresponding aroma release control instructions for the various fragrance base notes matched according to the real-time period, and the cycle duration is less than or equal to the real-time period.

4. The intelligent audio-driven fragrance control system according to claim 2, wherein The user interaction module further includes a music element volume adjustment unit, which is configured to manually adjust the volume values of one or more music elements; According to the adjustment operation of the music element volume adjustment unit, the AI audio driving module updates the release concentration parameter in the aroma release control instruction.

5. The intelligent audio-driven fragrance control system according to claim 2, wherein The user interaction module further includes an aroma adjustment unit, which is configured to manually adjust the current types of fragrance base notes, release concentration, release frequency, and / or release power; According to the adjustment operation of the aroma adjustment unit, the controller adjusts the valve opening degree, valve switching frequency, and / or valve switching speed of the corresponding aroma container.

6. The intelligent audio-driven fragrance control system according to claim 5, characterized in that, According to the adjustment operation of the aroma adjustment unit, the AI audio driving module analyzes the user's preferences and optimizes the matching rule between the type of music element and the fragrance base note and / or the relationship between the signal characteristics of the music element and the aroma release control instruction.

7. The intelligent audio-driven fragrance control system according to claim 1, wherein The system further includes an environmental sensor configured to detect one or more types of information such as temperature and humidity; The AI audio driving module is configured to generate the fragrance release control instruction in combination with the detection result of the environmental sensor.

8. The intelligent audio-driven fragrance control system according to claim 1, characterized in that, The fragrance container is configured with a driving unit for driving the change of the orientation of the fragrance release port, and the driving unit is connected to the controller; The AI audio driving module is configured to generate a control instruction for the orientation of the release port according to the signal characteristics of the music elements, and the controller controls the driving unit to act according to the control instruction for the orientation of the release port.

9. The intelligent audio-driven fragrance control system according to any one of claims 2 to 6, characterized in that, The user interaction module is an in-vehicle intelligent operation device or an application client on a smart mobile terminal.

10. The intelligent audio-driven fragrance control system according to any one of claims 1 to 8, characterized in that The AI audio driving module decomposes the real-time audio into a time series set of a plurality of music elements in the following manner: Collect the separated waveforms of multiple types of musical instruments and the full waveform of the mixed audio, and manually label classification labels for the separated waveforms and the full waveform respectively as a learning sample set for model training and verification; Input the learning sample set into an initial model based on a convolutional neural network and a long short-term memory neural network, learn to extract the time-frequency characteristics of the musical instruments, and implement multi-object learning of the initial model through the time-frequency masking method to obtain an encoder-decoder model capable of predicting the separated waveforms; Use the encoder-decoder model to predict the separated waveforms corresponding to a plurality of single sound sources arranged in time sequence in the real-time audio; Identify the sound source attributes matched by the separated waveforms, and the sound source attributes include one or more of loudness, frequency, and pitch; According to the separated waveforms of each single sound source obtained by parsing, and in combination with the sound source attributes matched thereto, obtain each decomposed music element.

11. The intelligent audio-driven fragrance control system according to claim 10, characterized in that, The initial model is configured with a plurality of CNN layers and a plurality of LSTM layers. The CNN layers are used to extract the characteristics of each section in the input real-time audio signal, generate the CNN characteristics of each section, and perform spatial modeling on each section in the input signal based on the CNN characteristics; Use the LSTM layers to process the CNN characteristics to generate the LSTM characteristics of each section, and perform temporal modeling on the CNN characteristics and each section based on the LSTM characteristics; The initial model is jointly trained based on the information after spatial modeling and temporal modeling to extract the time-frequency characteristics of the musical instruments to generate the encoder-decoder model.

12. An intelligent audio-driven fragrance control method, characterized in that, Including the following steps: Receive a real-time audio signal; Decompose the real-time audio into a time series set of a plurality of music elements by using a neural network model; Determine the types and signal characteristics of each music element, and the signal characteristics include one or more of volume, frequency, and rhythm information; According to the preset matching rules, determine the fragrance base tone matched by the type of the music element; and generate a fragrance release control instruction according to the signal characteristics of the music element, and the instruction includes one or more of release concentration, release frequency, and release power; Determine the target fragrance container according to the matched fragrance base tone, and control one or more of the valve opening degree, valve switching frequency, and valve switching speed of the fragrance release port corresponding to the target fragrance container according to the generated fragrance release control instruction.

13. A vehicle-mounted system, characterized in that, It includes a vehicle, an in-vehicle audio system provided inside the vehicle, and the intelligent audio-driven fragrance control system according to any one of claims 1 to 11, wherein the in-vehicle audio system is configured to play the real-time audio signal received by the audio input module of the intelligent audio-driven fragrance control system.

14. The vehicle-mounted system according to claim 13, wherein The AI audio driving module of the intelligent audio-driven fragrance control system is further configured to: Obtain the internal space information of the vehicle and determine the location information of the fragrance device in the internal space of the vehicle; Generate a fragrance release control instruction according to the location information of the fragrance device.

Images