Vehicle atmosphere adjusting method and system, storage medium and equipment

Abstract

The invention provides a vehicle atmosphere adjusting method and system, a storage medium and equipment, and relates to the technical field of intelligent driving, and the method comprises the steps: obtaining an environment sound signal and vehicle exterior video data; identifying acoustic features of each sound type included in the environment sound signal by using a Mel spectrum; extracting scene features of the video data outside the vehicle, and determining scene semantic features corresponding to the scene features; determining a vehicle environment style type according to the acoustic features and the scene semantic features; the vehicle environment style type is used for indicating and adjusting at least one of a vehicle-mounted sound system, a vehicle-mounted aromatherapy machine and a lighting system in a compartment. According to the invention, the internal atmosphere of the vehicle can be adjusted, the power of the vehicle-mounted equipment can be adjusted, the energy consumption is reduced, the intelligent level of the vehicle is improved, and a more comfortable and pleasant driving and riding environment is provided.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and in particular to a vehicle atmosphere adjustment method, system, storage medium and device. Background Technology

[0002] Currently, the interior controls rely on manual adjustments by the user, making them unable to respond to complex environmental changes. Furthermore, functions such as music recommendation, fragrance control, and lighting adjustment are independent of each other, resulting in a poor user experience.

[0003] Therefore, how to improve the intelligence level and user experience of in-vehicle systems is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide a vehicle atmosphere adjustment method, system, computer-readable storage medium, and electronic device that can dynamically adjust the interior atmosphere of a vehicle based on the vehicle environment to improve the user's driving experience.

[0005] To solve the above-mentioned technical problems, this application provides a vehicle atmosphere adjustment method, the specific technical solution of which is as follows:

[0006] Acquire ambient sound signals and external video data;

[0007] The acoustic features of each sound type contained in the environmental sound signal are identified using Mel spectrum analysis.

[0008] Extract scene features from the external video data and determine the scene semantic features corresponding to the scene features;

[0009] The vehicle environment style type is determined based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of the in-vehicle audio, in-vehicle aroma diffuser and interior lighting system.

[0010] Optionally, before using Mel spectrum to identify the acoustic features of each sound type contained in the ambient sound signal, the method further includes:

[0011] A number of triangular filters are set at equal intervals along the frequency, and the logarithm of the output of each triangular filter is taken to obtain the logarithmic Mel spectrum.

[0012] The logarithmic Mel spectrum is subjected to discrete cosine transformation, and the first few coefficients are taken as Mel frequency cepstral coefficients.

[0013] Optionally, using Mel spectrum to identify the acoustic features of each sound type contained in the environmental sound signal includes:

[0014] The ambient sound signal is divided into several short frames;

[0015] Apply the target window function to each of the short frames;

[0016] Perform a fast Fourier transform on each of the short frames to obtain the frequency domain amplitude spectrum;

[0017] By comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the environmental sound signal and the acoustic features corresponding to the sound type are determined.

[0018] Optionally, by comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the ambient sound signal and the acoustic features corresponding to the sound type are determined, including:

[0019] By comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the environmental sound signal and the acoustic characteristics corresponding to the sound type are determined based on the energy peak at different frequencies.

[0020] Alternatively, the log-Mel spectrum and the frequency domain amplitude spectrum can be input into a preset machine learning model to obtain the sound type output by the preset machine learning model and the acoustic features corresponding to the sound type.

[0021] Optionally, extracting scene features from the vehicle exterior video data and determining the scene semantic features corresponding to the scene features includes:

[0022] Extract street features and vegetation features from the vehicle exterior video data;

[0023] The scene type is determined based on the street features and the vegetation features;

[0024] The scene semantic database is invoked to determine the scene semantic features corresponding to the scene type.

[0025] Optionally, determining the vehicle environment style type based on the acoustic features and the scene semantic features includes:

[0026] The acoustic features and the scene semantic features are weighted and fused according to the set attention weights to obtain the fused features;

[0027] The fused features are input into the classification layer, and an activation function is used to output the probability distribution of the environment style type.

[0028] The label with the highest probability is selected from the probability distribution as the vehicle environment style type.

[0029] Optionally, after determining the vehicle environment style type based on the acoustic features and the scene semantic features, the method further includes:

[0030] Play the playlist music corresponding to the vehicle's ambient style type using the car audio system;

[0031] The multi-channel micro-pump is used to control the car air freshener to adjust the evaporation concentration of the fragrance capsules;

[0032] The interior lighting system is activated to adjust the hue, saturation, and flashing frequency of the LEDs.

[0033] This application also provides a vehicle atmosphere adjustment system, including:

[0034] The data acquisition module is used to acquire ambient sound signals and external video data.

[0035] An acoustic feature extraction module is used to identify the acoustic features of each sound type contained in the environmental sound signal using Mel spectrum.

[0036] The scene feature extraction module is used to extract scene features from the vehicle exterior video data and determine the scene semantic features corresponding to the scene features;

[0037] The style type determination module is used to determine the vehicle environment style type based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of the in-vehicle audio system, in-vehicle aroma diffuser and interior lighting system.

[0038] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0039] This application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described above when it invokes the computer program in the memory.

[0040] This application provides a vehicle atmosphere adjustment method, comprising: acquiring ambient sound signals and external video data; using Mel spectrum to identify the acoustic features of each sound type contained in the ambient sound signals; extracting scene features from the external video data and determining scene semantic features corresponding to the scene features; determining a vehicle environment style type based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of a car audio system, a car air freshener, and an interior lighting system.

[0041] This application acquires environmental sound signals and external video data, and further extracts acoustic and scene features to comprehensively determine the style type of the environment in which the vehicle is located. It can further adjust the in-vehicle audio system, in-vehicle aroma diffuser, and interior lighting system to achieve interior atmosphere adjustment. It can also reasonably adjust the power of in-vehicle equipment, reduce energy consumption, improve the vehicle's intelligence level, and provide a more comfortable and pleasant driving environment.

[0042] This application also provides a vehicle atmosphere control system, a computer-readable storage medium, and an electronic device, which have the aforementioned beneficial effects, and will not be elaborated here. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0044] Figure 1 A flowchart illustrating a vehicle atmosphere adjustment method provided in this application embodiment;

[0045] Figure 2 This is a schematic diagram of a vehicle atmosphere control system provided in an embodiment of this application;

[0046] Figure 3 This is a structural diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] See Figure 1 , Figure 1 A flowchart of a vehicle atmosphere adjustment method provided in this application embodiment, the method including:

[0049] S101: Acquire ambient sound signals and external video data;

[0050] S102: Identify the acoustic features of each sound type contained in the environmental sound signal using Mel spectrum;

[0051] S103: Extract scene features from the external video data and determine the scene semantic features corresponding to the scene features;

[0052] S104: Determine the vehicle environment style type based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of the in-vehicle audio system, in-vehicle aroma diffuser and interior lighting system.

[0053] Environmental sound signals and external video data can be acquired independently or simultaneously. If acquired independently, the two types of information do not interfere with each other. If acquired simultaneously, the identical bitstream facilitates subsequent feature recognition, particularly fusion feature extraction.

[0054] There is no limitation on the type of ambient sound signal acquired; a high-sensitivity microphone array can be used to collect ambient sound signals from outside the vehicle. The microphone array can be installed at the front, sides, and rear of the vehicle to cover different directions. In one feasible implementation, the sampling rate of the microphone can be ensured to be no less than 16kHz to capture sufficient audio details. In addition, the acquired ambient sound signals can be preliminarily processed to remove background noise and interference signals, thereby improving the accuracy of acoustic feature identification in step S102.

[0055] When collecting video data outside the vehicle, a high-definition camera can be used. The camera can be mounted on the windshield, rearview mirror, or rear of the vehicle. In one feasible implementation, the video frame rate should be at least 15fps and the resolution at least 720p to ensure video data quality. Similarly, the collected video data can be preliminarily processed to remove blurry and distorted parts.

[0056] In step S102, it is necessary to use the Mel spectrum to identify the acoustic characteristics of each sound type contained in the environmental sound signal. Before this, several triangular filters can be set at equal intervals along the frequency. The logarithm of the output of each triangular filter is taken to obtain the logarithmic Mel spectrum. Then, the logarithmic Mel spectrum is subjected to discrete cosine transformation, and the first few coefficients are taken as Mel frequency cepstral coefficients.

[0057] Using a triangular filter to divide frequency bands according to the Mel scale can be used to mimic the nonlinear perception of frequency by the basilar membrane of the human ear (high resolution for low frequencies, low resolution for high frequencies), converting physical frequencies into a representation that is closer to human subjective auditory perception. By taking the logarithm of the filtered energy value, the numerical dynamic range (such as the difference between loud and soft sounds) can be significantly compressed, absolute intensity differences can be weakened, relative energy distribution can be strengthened, and multiplicative noise (such as background noise in ambient sound signals) can be suppressed. Since the cepstral domain is naturally good at separating the periodic components (fundamental frequency F0) and their harmonic structure in a signal, and effectively separating the articulation mode (such as voiced and unvoiced sounds) and vocal tract resonance characteristics, the Mel frequency cepstral coefficients can ensure that the extracted acoustic features are more robust.

[0058] Then, when executing step S102, the following steps may be included:

[0059] The first step is to divide the ambient sound signal into several short frames;

[0060] The second step is to apply a target window function to each of the short frames.

[0061] The third step is to perform a fast Fourier transform on each of the short frames to obtain the frequency domain amplitude spectrum;

[0062] The fourth step is to compare the frequency domain amplitude spectrum and the log-Mel spectrum to determine the sound type contained in the environmental sound signal and the acoustic features corresponding to the sound type.

[0063] Since ambient sound signals are usually continuous, and short-time signals are easier to process, segmenting long signals into short frames can better capture the local and key features of ambient sound signals. For example, a continuous ambient sound signal can be divided into multiple short frames at certain time intervals. Each short frame contains a small segment of sound signal.

[0064] Window functions are used to reduce signal discontinuities at frame edges and prevent spectral leakage. The target window function is not limited here and can include Hamming windows, Hanning windows, etc. When applying the target window function to each short frame, each short frame can be multiplied by a window function.

[0065] Subsequently, a Fast Fourier Transform (FFT) is performed on each short frame to obtain its frequency domain representation. The frequency domain amplitude spectrum can be obtained by taking the magnitude of the FFT result, i.e., |X(k)|, where X(k) is the result of the FFT.

[0066] Finally, the frequency domain amplitude spectrum is converted to a Mel spectrum, and then the logarithm is taken. Calculating the Mel spectrum typically involves mapping frequency segments to Mel frequencies and summing the energy for each band. Features, such as Mel frequency cepstral coefficients (MFCCs), are extracted from the logarithmic Mel spectrum.

[0067] At this point, the frequency domain amplitude spectrum and the log-Mel spectrum can be compared to determine the sound type contained in the environmental sound signal and the corresponding acoustic features based on the energy peaks at different frequencies. Alternatively, the log-Mel spectrum and the frequency domain amplitude spectrum can be input into a preset machine learning model to obtain the sound type output by the preset machine learning model and the corresponding acoustic features. The choice of preset machine learning model is not limited here; for example, a convolutional neural network model may be used.

[0068] In step S103, street features and vegetation features can be extracted from the video data outside the vehicle, and then the scene type can be determined based on the street features and vegetation features, thereby calling the scene semantic database to determine the scene semantic features corresponding to the scene type.

[0069] Extract street features (such as roads, buildings, traffic signs, etc.) and vegetation features (such as trees, grass, flowers, etc.) from the exterior video data. Use pre-trained convolutional neural networks such as ResNet to extract visual features from each frame of the image. Perform dimensionality reduction on the extracted visual features to facilitate subsequent analysis. Based on the extracted street and vegetation features, determine the scene type (such as urban streets, rural roads, natural environments, etc.). Access a scene semantic database to find scene semantic features corresponding to the scene type, such as the busyness of an urban environment or the tranquility of a natural environment.

[0070] In step S104, the aim is to determine the vehicle environment style type based on the acoustic features and the scene semantic features. The environment style type can be predefined, and corresponding music style, aroma release strategy and interior lighting configuration scheme can be adapted to the environment style type.

[0071] One feasible implementation may include the following steps:

[0072] S1041: The acoustic features and the scene semantic features are weighted and fused according to the set attention weights to obtain fused features;

[0073] S1042: Input the fused features into the classification layer and use an activation function to output the probability distribution of the environment style type;

[0074] S1043: Select the label with the highest probability from the probability distribution as the vehicle environment style type.

[0075] Acoustic features and scene semantic features can be weighted and fused according to the set attention weights to obtain fused features. The attention weights can be adjusted according to actual needs to highlight the importance of audio or video features. The fused features are input into the classification layer, and an activation function (such as softmax) is used to output the probability distribution of the environment style type. Then, the label with the highest probability is selected as the vehicle environment style type.

[0076] When adjusting the vehicle's ambiance, at least one of the following can be adjusted: the car audio system, the car air freshener, and the interior lighting system. Usually, all three systems need to be adjusted simultaneously to suit the desired vehicle environment style.

[0077] For car audio systems, adjustments are made based on the ambient sound style, playing the corresponding playlist. For example, birdsong or a babbling brook can be played in a "natural tranquility" environment, while soothing music can be played in a "busy city" environment. The volume and sound effects can also be adjusted to suit different environments. Specifically, you can directly access the playlist corresponding to the vehicle's ambient sound style.

[0078] For car air fresheners, a multi-channel micro-pump can be used to adjust the concentration of fragrance capsules based on the style of the environment. For example, a refreshing herbal scent can be released in a "natural and tranquil" environment, while a soothing lavender scent can be released in a "busy city" environment. In practical applications, the multi-channel micro-pump can be used to control the car air freshener to adjust the concentration and release frequency of the fragrance capsules to provide the best olfactory experience.

[0079] For the interior lighting system, the hue, saturation, and flashing frequency of the LEDs can be adjusted. Depending on the ambient style, the hue, saturation, and flashing frequency of the LEDs can be adjusted using the vehicle's lighting system. For example, soft, warm-toned lighting can be used in a "natural and tranquil" environment, while bright, cool-toned lighting can be used in a "busy urban" environment. Adjust the brightness and flashing mode of the lights to create an atmosphere that matches the style of the environment.

[0080] This application embodiment acquires environmental sound signals and external video data, and further extracts acoustic features and scene features to comprehensively determine the style type of the environment in which the vehicle is located. It can further adjust the in-vehicle audio, in-vehicle aroma diffuser and interior lighting system to achieve the adjustment of the vehicle's interior atmosphere. It can reasonably adjust the power of in-vehicle equipment, reduce energy consumption, improve the vehicle's intelligence level, and provide a more comfortable and pleasant driving environment.

[0081] In one feasible implementation, acoustic features and scene semantic features can also be used as a fusion process, as follows:

[0082] Step 1: Acquire ambient sound signals and external video data:

[0083] Sampling rate standardization: unifying the sampling rate of audio signals to a standard value.

[0084] Frame segmentation and windowing: The audio signal is divided into short frames, each typically 20-40ms long, with a frame shift of about 10ms. A window function (such as a Hamming window) is applied to each frame to reduce spectral leakage.

[0085] The acoustic features of each sound type contained in the environmental sound signal are identified using Mel spectrum.

[0086] Read video frames: Use video processing tools (such as OpenCV) to read each frame of the video. If the video frame rate is too high, downsampling can be performed to reduce computation. Additionally, images can be cropped or scaled as needed to remove irrelevant regions or standardize image size. Use a pre-trained CNN (such as ResNet or VGG) to extract scene features from the vehicle exterior video data and determine the corresponding scene semantic features.

[0087] Step 2: Construct a two-stream neural network.

[0088] Audio streaming networks include:

[0089] Feature extraction layer: The extracted audio features are input into the audio stream network. Convolutional neural networks (CNNs) or recurrent neural networks (RNNs) can be used to process the audio features.

[0090] Feature encoding layer: Encodes audio features through multiple convolutional or recurrent layers to extract higher-level feature representations.

[0091] Video streaming networks include:

[0092] Feature extraction layer: Inputs the extracted visual features into the video stream network. Pre-trained CNNs (such as ResNet and VGG) can be used as feature extractors.

[0093] Feature encoding layer: Encodes visual features through multiple convolutional or recurrent layers to extract higher-level feature representations.

[0094] During the attention fusion phase, ensure that the features of the audio and video streams are aligned in the temporal dimension. For example, if audio features are extracted once per second, video features should also be extracted once per second. Simultaneously, apply self-attention mechanisms to both the audio and video streams to enhance the representational power of the features. Self-attention mechanisms help the model better focus on important feature components.

[0095] Furthermore, a cross-attention mechanism can be applied between the audio and video streams to achieve feature fusion. The cross-attention mechanism helps the model learn the correlation between audio and video features.

[0096] The audio and video features are weighted and fused based on attention weights. These attention weights can be dynamically calculated or fixed values.

[0097] Fully connected layer: The fused features are input into the fully connected layer for further feature extraction. The fused features are then input into the classification layer, where the softmax function outputs the probability distribution of the environment style labels. The classification layer can be a fully connected layer, with its output dimension matching the number of environment style labels.

[0098] Step 3: Determine the vehicle environment style type based on the acoustic features and the scene semantic features: Select the label with the highest probability as the final environment style label based on the probability distribution.

[0099] When preparing training data, you can prepare labeled audio and video data to ensure the diversity and representativeness of the data. Divide the data into training set, validation set, and test set.

[0100] During model training, a two-stream neural network is trained using labeled data. The classification loss function (such as cross-entropy loss) is optimized during training. Hyperparameters are tuned using a validation set to ensure the model's generalization ability.

[0101] Afterwards, model evaluation can be performed to assess the model's performance on the test set, using metrics such as accuracy, recall, and F1 score to measure the model's classification effectiveness.

[0102] After the model passes evaluation, the trained model will be deployed to practical applications, such as real-time environmental monitoring systems, to ensure that the model's inference speed meets the needs of practical applications.

[0103] In real-time applications, the input audio and video streams are preprocessed, feature extracted, fused, and classified. The output is an environmental style label for the vehicle's environmental style type, providing users with real-time environmental information.

[0104] Through the above steps, a system combining a two-stream neural network and an attention fusion mechanism can be implemented to take acoustic and visual features as input and output environmental style labels. This method can fully utilize audio and video information to improve the accuracy and robustness of environmental style recognition, thus adjusting the vehicle's atmosphere.

[0105] See Figure 2 , Figure 2 This is a schematic diagram of a vehicle atmosphere adjustment system provided in an embodiment of this application. The system includes:

[0106] The data acquisition module is used to acquire ambient sound signals and external video data.

[0107] An acoustic feature extraction module is used to identify the acoustic features of each sound type contained in the environmental sound signal using Mel spectrum.

[0108] The scene feature extraction module is used to extract scene features from the vehicle exterior video data and determine the scene semantic features corresponding to the scene features;

[0109] The style type determination module is used to determine the vehicle environment style type based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of the in-vehicle audio system, in-vehicle aroma diffuser and interior lighting system.

[0110] Based on the above embodiments, as a preferred embodiment, it further includes:

[0111] The Mel spectrum acquisition module is used to set several triangular filters at equal intervals along the frequency, take the logarithm of the output of each triangular filter to obtain the logarithmic Mel spectrum, and perform discrete cosine transformation on the logarithmic Mel spectrum, taking the first few coefficients as Mel frequency cepstral coefficients.

[0112] Based on the above embodiments, as a preferred embodiment, the acoustic feature extraction module includes:

[0113] A signal segmentation unit is used to segment the ambient sound signal into several short frames;

[0114] A window function application unit is used to apply a target window function to each of the short frames;

[0115] The time-frequency conversion unit is used to perform a fast Fourier transform on each of the short frames to obtain the frequency domain amplitude spectrum;

[0116] The feature recognition unit is used to compare the frequency domain amplitude spectrum and the log-Mel spectrum to determine the sound type contained in the environmental sound signal and the acoustic features corresponding to the sound type.

[0117] Based on the above embodiments, as a preferred embodiment, the feature recognition module is a unit for performing the following steps:

[0118] By comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the environmental sound signal and the acoustic features corresponding to the sound type are determined based on the energy peak at different frequencies; or, the log-Mel spectrum and the frequency domain amplitude spectrum are input into a preset machine learning model to obtain the sound type output by the preset machine learning model and the acoustic features corresponding to the sound type.

[0119] Based on the above embodiments, as a preferred embodiment, the scene feature extraction module includes:

[0120] An environmental feature extraction unit is used to extract street features and vegetation features from the vehicle exterior video data;

[0121] A scene recognition unit is used to determine the scene type based on the street features and the vegetation features;

[0122] The scene feature definition unit is used to call the scene semantic database to determine the scene semantic features corresponding to the scene type.

[0123] Based on the above embodiments, as a preferred embodiment, the style type determination module includes:

[0124] The feature fusion unit is used to weight and fuse the acoustic features and the scene semantic features according to a set attention weight to obtain fused features;

[0125] The probability distribution acquisition unit is used to input the fused features into the classification layer and output the probability distribution of the environment style type using an activation function;

[0126] The type determination unit is used to select the label with the highest probability from the probability distribution as the vehicle environment style type.

[0127] Based on the above embodiments, as a preferred embodiment, it further includes:

[0128] The in-vehicle atmosphere adjustment module is used to perform the following steps: call the in-vehicle audio system to play the playlist music corresponding to the vehicle's environmental style type; call the multi-channel micro-pump to control the in-vehicle aroma diffuser to adjust the evaporation concentration of the fragrance capsules; and call the in-vehicle lighting system to adjust the hue, saturation, and flashing frequency of the light-emitting diodes.

[0129] This application also provides an embodiment corresponding to a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method described in the above method embodiments.

[0130] It is understood that if the methods in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and executes all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0131] The computer-readable storage medium provided in this embodiment includes the method mentioned above, and has the same effect.

[0132] This application also provides an electronic device, see [link to document]. Figure 3 The present application provides a structural diagram of an electronic device, such as... Figure 3 As shown, it may include a processor 1410 and a memory 1420.

[0133] The processor 1410 may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor 1410 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 1410 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 1410 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor 1410 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0134] The memory 1420 may include one or more computer-readable storage media, which may be non-transitory. The memory 1420 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In this embodiment, the memory 1420 is used to store at least the following computer program 1421, which, after being loaded and executed by the processor 1410, is capable of implementing the relevant steps in the methods executed by the electronic device side as disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 1420 may also include an operating system 1422 and data 1423, etc., and the storage method may be temporary storage or permanent storage. The operating system 1422 may include Windows, Linux, Android, etc.

[0135] In some embodiments, the electronic device may further include a display screen 1430, an input / output interface 1440, a communication interface 1450, a sensor 1460, a power supply 1470, and a communication bus 1480.

[0136] certainly, Figure 3 The structure of the electronic device shown does not constitute a limitation on the electronic device in the embodiments of this application. In practical applications, the electronic device may include more than [other components]. Figure 3 More or fewer components as shown, or combinations of certain components.

[0137] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. As the system provided in the embodiments corresponds to the method provided in the embodiments, the description is relatively simple; relevant parts can be found in the method section.

[0138] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

[0139] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A method for adjusting the atmosphere of a vehicle, characterized in that, include: Acquire ambient sound signals and external video data; The acoustic features of each sound type contained in the environmental sound signal are identified using the Mel spectrum. Extract scene features from the external video data and determine the scene semantic features corresponding to the scene features; The vehicle environment style type is determined based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of the in-vehicle audio, in-vehicle aroma diffuser and interior lighting system.

2. The vehicle atmosphere adjustment method according to claim 1, characterized in that, Before using Mel spectrum to identify the acoustic features of each sound type contained in the environmental sound signal, the process further includes: A number of triangular filters are set at equal intervals along the frequency, and the logarithm of the output of each triangular filter is taken to obtain the logarithmic Mel spectrum. The logarithmic Mel spectrum is subjected to discrete cosine transformation, and the first few coefficients are taken as Mel frequency cepstral coefficients.

3. The vehicle atmosphere adjustment method according to claim 2, characterized in that, The acoustic features used to identify the various sound types contained in the environmental sound signal using Mel spectrum include: The ambient sound signal is divided into several short frames; Apply the target window function to each of the short frames; Perform a fast Fourier transform on each of the short frames to obtain the frequency domain amplitude spectrum; By comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the environmental sound signal and the acoustic features corresponding to the sound type are determined.

4. The vehicle atmosphere adjustment method according to claim 3, characterized in that, By comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the environmental sound signal and the acoustic features corresponding to the sound type are determined, including: By comparing the frequency domain amplitude spectrum and the log-Mel spectrum, the sound type contained in the environmental sound signal and the acoustic characteristics corresponding to the sound type are determined based on the energy peak at different frequencies. Alternatively, the log-Mel spectrum and the frequency domain amplitude spectrum can be input into a preset machine learning model to obtain the sound type output by the preset machine learning model and the acoustic features corresponding to the sound type.

5. The vehicle atmosphere adjustment method according to claim 1, characterized in that, Extracting scene features from the vehicle exterior video data and determining the corresponding scene semantic features includes: Extract street features and vegetation features from the vehicle exterior video data; The scene type is determined based on the street features and the vegetation features; The scene semantic database is invoked to determine the scene semantic features corresponding to the scene type.

6. The vehicle atmosphere adjustment method according to claim 1, characterized in that, Determining the vehicle environment style type based on the acoustic features and the scene semantic features includes: The acoustic features and the scene semantic features are weighted and fused according to the set attention weights to obtain the fused features; The fused features are input into the classification layer, and an activation function is used to output the probability distribution of the environment style type. The label with the highest probability is selected from the probability distribution as the vehicle environment style type.

7. The vehicle atmosphere adjustment method according to claim 1, characterized in that, After determining the vehicle environment style type based on the acoustic features and the scene semantic features, the method further includes: Play the playlist music corresponding to the vehicle's ambient style type using the car audio system; The multi-channel micro-pump is used to control the car air freshener to adjust the evaporation concentration of the fragrance capsules; The interior lighting system is activated to adjust the hue, saturation, and flashing frequency of the LEDs.

8. A vehicle atmosphere adjustment system, characterized in that, include: The data acquisition module is used to acquire ambient sound signals and external video data. An acoustic feature extraction module is used to identify the acoustic features of each sound type contained in the environmental sound signal using Mel spectrum. The scene feature extraction module is used to extract scene features from the vehicle exterior video data and determine the scene semantic features corresponding to the scene features; The style type determination module is used to determine the vehicle environment style type based on the acoustic features and the scene semantic features; the vehicle environment style type is used to indicate the adjustment of at least one of the in-vehicle audio system, in-vehicle aroma diffuser and interior lighting system.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed, implements the steps of the method as described in any one of claims 1 to 7.

Images