Sound source switching method and device based on priority judgment, equipment and storage medium

Abstract

This application discloses a method, apparatus, device, and storage medium for sound source switching based on priority judgment, applied to a smart audio device. The method includes acquiring sound source device access information of the smart audio device; determining whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information; performing state machine matching on the sound source signals in the sound source selection agenda to obtain new sound source signals with event trigger tags; updating denoising rules according to a preset state machine, and denoising the new sound source signals with event trigger tags to generate sound source confirmation result information; and generating and sending a confirmation connection command for the sound source signal matching the smart audio device based on the sound source confirmation result information. The method of this application embodiment enables the system to have adaptive optimization capabilities for complex acoustic environments and personalized usage habits by establishing a multi-sound source parallel monitoring model and a global state matrix.

Description

Technical Field

[0001] This application relates to the field of intelligent playback device technology, and in particular to a sound source switching method, apparatus, device and storage medium based on priority judgment. Background Technology

[0002] When faced with complex application scenarios involving multiple audio sources such as HDMI, WiFi, and Bluetooth, existing smart audio devices generally still rely on traditional control logic such as "fixed input selection" or "passive trigger switching." This mechanism requires users to manually switch input channels physically via an app or remote control, or the device will only mechanically switch to the next preset input after a detected input signal is physically disconnected. Because this approach lacks parallel and continuous monitoring of multiple channel states and cannot effectively identify the content type of the audio stream, the device struggles to intelligently determine which audio source truly matches the user's intended playback source. In practical applications, typical problems include the TV continuing to occupy the audio source even when it's in sleep mode because the interface hasn't been physically disconnected, the device failing to automatically and seamlessly take over when a mobile phone initiates wireless playback, and random Bluetooth signal preemption. This technical deficiency not only forces users to perform frequent interactive operations, creating a fragmented auditory experience, but also significantly increases the learning curve and usage threshold for users due to inconsistencies in switching logic between different brands. Summary of the Invention

[0003] This application provides a sound source switching method, apparatus, device, and storage medium based on priority judgment, aiming to solve the problem of poor adaptability of existing audio devices in the process of accessing sound source signals.

[0004] In a first aspect, embodiments of this application provide a sound source switching method based on priority judgment, applied to a smart audio device. The method includes obtaining sound source device access information of the smart audio device; determining whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information; selecting sound source signals in the pre-access state and adding them to a preset multi-sound source parallel state monitoring model, wherein the smart audio device performs sound source connection priority ranking in the multi-sound source parallel state monitoring model; obtaining sound source state judgment information for each sound source signal; establishing a sound source state vector based on the sound source state judgment information; and assigning each sound source state... The state vectors are merged to generate a global state matrix of the sound source; sound source signals with connectable states are selected from the global state matrix and added to the sound source selection sequence, and the sound source signals are arranged in the sound source selection agenda in the sound source selection agenda; the sound source signals in the sound source selection agenda are matched with a state machine to obtain new sound source signals with event trigger tags; the denoising rules are updated according to the preset state machine, and the new sound source signals with event trigger tags are denoised to generate sound source confirmation result information; based on the sound source confirmation result information, a confirmation connection command is generated and sent for the sound source signal that matches the smart audio device.

[0005] Secondly, embodiments of this application also provide a sound source switching device based on priority judgment, comprising: a first judgment unit, used to acquire sound source device access information of a smart audio device, and determine whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information; a priority arrangement unit, used to select sound source signals in the pre-access state and add them to a preset multi-sound source parallel state monitoring model, wherein the smart audio device performs sound source connection priority arrangement in the multi-sound source parallel state monitoring model; a second judgment unit, used to acquire sound source state judgment information of each sound source signal, and establish a sound source state vector based on the sound source state judgment information; and a state matrix generation unit, used to generate a state matrix for each sound source... The device employs a system of merging state vectors to generate a global state matrix for the sound source. An agenda-arranging unit selects connectable sound source signals from the global state matrix and adds them to a sound source optimization sequence, where the sound source signals undergo an agenda-arranging process. A tagging unit performs state machine matching on the sound source signals in the agenda-arranging process to obtain new sound source signals with event-triggered tags. A denoising unit updates denoising rules according to a preset state machine and denoises the new sound source signals with event-triggered tags to generate sound source confirmation result information. An instruction generation unit generates and sends a confirmation connection instruction to the sound source signal matching the smart audio device based on the sound source confirmation result information. Furthermore, the device includes an information acquisition unit for sensing available signals from the smart audio device to obtain sound source device access information, and a determination unit for determining whether the state parameters in the sound source device access information are preset state values ​​to determine whether the smart audio device is in a pre-access state.

[0006] Thirdly, embodiments of this application also provide a computer device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0007] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, can implement the above-described method.

[0008] This application provides a method, apparatus, device, and storage medium for sound source switching based on priority judgment. The method includes: acquiring the sound source device access information of the smart audio device; determining whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information; selecting the sound source signals in the pre-access state and adding them to a preset multi-sound source parallel state monitoring model, in which the smart audio device prioritizes the sound source connections; acquiring the sound source state judgment information of each sound source signal and establishing a sound source state vector based on the sound source state judgment information; merging the sound source state vectors to generate a global sound source state matrix; selecting the connectable sound source signals in the global sound source state matrix and adding them to a sound source optimization sequence, in which the sound source signals are arranged in a sound source optimization agenda; performing state machine matching on the sound source signals in the sound source optimization agenda to obtain new sound source signals with event trigger tags; updating the denoising rules according to the preset state machine and denoising the new sound source signals with event trigger tags to generate sound source confirmation result information; and generating and sending a confirmation connection command for the sound source signals that match the smart audio device based on the sound source confirmation result information. The above method establishes a multi-source parallel monitoring model and a global state matrix to achieve dynamic weighted scoring decisions based on source activity, content type semantics, and signal quality. It effectively suppresses transient signal interference and frequent jumps by using hysteresis control rules and double-layer labeling denoising technology. Combined with user intent prediction and decoding resource preloading mechanism, it eliminates the cold start delay of the audio path to achieve seamless transition. Furthermore, by adjusting model parameters and user preference weights through real-time feedback, the system has the ability to adaptively optimize for complex acoustic environments and personalized usage habits. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A schematic diagram illustrating an application scenario of the sound source switching method based on priority judgment provided in this application embodiment; Figure 2 A flowchart illustrating the sound source switching method based on priority determination provided in an embodiment of this application; Figure 3 A schematic diagram of a sub-process of the sound source switching method based on priority determination provided in an embodiment of this application; Figure 4 A schematic block diagram of a sound source switching device based on priority determination provided in an embodiment of this application; Figure 5 A schematic block diagram of a computer device provided in an embodiment of this application. Detailed Implementation

[0011] 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.

[0012] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0013] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0014] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0015] This application provides a method, apparatus, device, and storage medium for sound source switching based on priority judgment.

[0016] The execution subject of the priority-based sound source switching method can be the priority-based sound source switching device provided in the embodiments of this application, or a computer device that integrates the priority-based sound source switching device. The priority-based sound source switching device can be implemented in hardware or software. The computer device can be a terminal or a server. The terminal can be a smartphone, tablet computer, handheld computer, or laptop computer, etc.

[0017] This priority-based sound source switching method is applied to Figure 5 Among the 500 computer devices.

[0018] Figure 1 This is a flowchart illustrating a sound source switching method based on priority determination provided in an embodiment of this application. The method includes the following steps S110-180.

[0019] S110. Obtain the sound source device access information of the smart audio device, and determine whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information.

[0020] S120. Select the sound source signal in the pre-access state and add it to the preset multi-source parallel state monitoring model. The intelligent audio device will arrange the sound source connection priority in the multi-source parallel state monitoring model.

[0021] S130. Obtain the sound source state judgment information of each sound source signal, and establish the sound source state vector based on the sound source state judgment information.

[0022] S140. Merge the state vectors of each sound source to generate a global state matrix of the sound sources.

[0023] S150. Select the connectable state of the sound source signal in the global state matrix of the sound source and add it to the sound source selection sequence. The sound source signal is arranged in the sound source selection sequence.

[0024] S160. Perform state machine matching on the sound source signals in the sound source selection agenda to obtain new sound source signals with event trigger labels.

[0025] S170. Update the denoising rules according to the preset state machine, and denoise the new sound source signal with the event trigger tag to generate sound source confirmation result information.

[0026] S180. Based on the sound source confirmation result information, generate and send a confirmation connection command for the sound source signal that matches the smart audio device.

[0027] For step S110, specifically, the system layer of the smart audio device (such as a sound bar) scans all physical interfaces and wireless protocol stacks through driver layer polling or interrupt monitoring. HDMI detects the HPD (Hot-plug Detection) pin level and CEC (Consumer Electronics Control) logical address. WiFi monitors the LAN SSDP discovery protocol, mDNS service, or AirPlay / Chromecast handshake request. Bluetooth monitors ACL link establishment requests and A2DP pairing status. The system defines the state where a signal is detected but a stable audio stream transmission has not yet been formally established as a "pre-access state." For example, HDMI has handshaked but no audio packet, Bluetooth has paired but no playback, and WiFi has connected but no screen mirroring. Establishing a sound source pool allows the system to perceive potential sound sources in advance, laying the foundation for subsequent parallel monitoring and rapid response, and avoiding cold start delays when the device actually initiates playback.

[0028] In step S120, the sound sources in the pre-access state are registered to the system's "multi-source parallel state monitoring model." The system loads the default basic priority configuration and initially sorts the sound sources in the current model according to preset rules (e.g., $P_{HDMI}>P_{WiFi}>P_{BT}$). At this time, the priority is mainly based on the interface type and the user's fixed preferences. This achieves parallel management of multiple channels, breaking the limitations of traditional single-point monitoring. By preset priorities, it ensures that, under normal circumstances, the system conforms to common habits (e.g., TV sound usually takes precedence over mobile phone sound), establishing a baseline for decision-making.

[0029] In step S130, the system continuously collects in-depth operational parameters for each sound source. These parameters are no longer limited to "connection / disconnection" but delve into the data link and content layers, and the link and physical layers: RSSI (Bluetooth signal strength), WiFi bandwidth / packet loss rate, HDMI EDID / clock stability, audio packet arrival rate, buffer level, audio format, and CEC power status commands. The system encapsulates this multi-dimensional information into a standardized sound source state vector $V_i(t)$, for example: $V_{hdmi} ={Link:Connected, Active:Silent, Content:Movie, Quality:High}$. This transforms heterogeneous sound source information (physical, network, and protocol-based) into a unified mathematical model (vector) that can be processed by a computer. This is a prerequisite for achieving fine-grained control, enabling subsequent scoring algorithms to make judgments based on specific signal quality rather than simple on / off states.

[0030] For step S140, within the same time slice $t$, the state vectors $V_{hdmi}(t), V_{wifi}(t), V_{bt}(t)$ of all sound sources are aggregated. A global state matrix $M(t)$ is constructed, with the horizontal axis representing the sound source dimension and the vertical axis representing the state attribute dimension. This matrix allows the system to view the real-time competitive situation of all sound sources, providing a global perspective for decision-making. The system no longer views a single sound source in isolation but can horizontally compare the states of different sound sources (e.g., comparing the silent state of HDMI with the active state of WiFi), providing a complete data panorama for subsequent optimal switching.

[0031] For step S150, sound sources with broken links or signal quality below the survival threshold are removed from the matrix, and "connectable" candidates are selected to join the "sound source selection sequence". Dynamic scoring and ranking (the core of the agenda): This is the most critical decision-making step. The system uses a preset weighted scoring function to score each sound source in the sequence: $$Score_i = w_p \cdot P_i+ w_a \cdot A_i + w_c \cdot C_i + w_q \cdot Q_i + w_u \cdot UserPref_i$$, introducing content type recognition. If it is identified as "system prompt tone" or "call", the weight of $C_i$ is adjusted; a user preference learning model $UserPref_i$ is introduced to fine-tune the weights. Based on the score of $Score_i$, the selection sequence is reordered, realizing "intent-based intelligent decision-making". Compared to fixed priority, this step can effectively prevent low-priority prompts from accidentally interrupting high-priority video playback. At the same time, it can dynamically adjust according to user habits (such as increasing WiFi priority when users are used to watching videos and listening to music at the same time), solving the problems of accidental triggering and rigid experience.

[0032] In another embodiment, the sound source signal is arranged in the sound source selection agenda within the sound source selection sequence by: an adaptive sound source arbitration algorithm based on a competitiveness index and time memory decay, wherein the calculation formula of the adaptive sound source arbitration algorithm is as follows: , in, Let be the competitive priority index of sound source i at time t. For the kth feature dimension of sound source i, the original value (activity, quality, content type, etc.) is used. The sigmoid normalization function is used for non-linear feature mapping. ; The dynamic weight of the k-th feature is adaptively adjusted over time and in different scenarios. The timestamp of the last active time for sound source i; The duration of silence for sound source i; λ is the memory decay coefficient (typical value 0.1-0.5), which controls the rate of forgetting of historical activity; The cumulative duration of continuous and stable playback of sound source i; τ is the stability time constant (e.g., 5 seconds), used for normalization; μ is the stability gain coefficient (e.g., 0.3), which rewards sound sources that are stable for a long time; The switching cost is the cost of switching from the current sound source j to the candidate sound source i. ; , Encode the audio source type (HDMI=3, WiFi=2, Bluetooth=1); For time window The cumulative number of switching times within (e.g., 60 seconds); For sound source i, the signal-to-noise ratio or signal quality index; , and These are the weighting coefficients for each cost component; η is the handover penalty coefficient (e.g., 0.2), which suppresses frequent handovers.

[0033] The following example illustrates the above solution in detail: A combined scenario of watching TV via HDMI and streaming music via mobile phone Wi-Fi. Current audio source: HDMI TV (j), playing for 30 minutes; New audio source: WiFi Music (i), just started playing for 2 seconds; Calculating the CPI of a WiFi sound source specifically includes: Step 1: Basic Scoring Section; Assume the normalized eigenvalues ​​are: Activity level σ(X1,WiFi) = σ(0.9) ≈ 0.85 Signal quality σ(X2,WiFi) = σ(0.95) ≈ 0.88 Content type σ(X3,WiFi) = σ(0.8) ≈ 0.75 Weighted sum: 0.4 × 0.85 + 0.3 × 0.88 + 0.3 × 0.75 = 0.829 Step 2: Memory decay factor; WiFi has just been activated; silent period = 0 seconds.

[0034] Step 3: Stability gain; The WiFi playback only lasts for 2 seconds, and the stable playback duration is relatively short.

[0035] Step 4: Switching costs; Type difference: 0.5 × |2 - 3| = 0.5 Recent handover count: 0.3 × 0 = 0 (assuming no frequent handovers) Signal quality: 0.2 × 1 / 0.95 ≈ 0.21 Total cost:

[0036] Step 5: Final CPI Calculation: . Simultaneously calculate the CPI of HDMI (assuming the calculation is as follows). =0.65): (0.758)> (0.65), and the difference exceeds the hysteresis threshold, so the system determines that it should switch to WiFi music.

[0037] For step S160, the system internally maintains multiple finite state machines (FSMs) for different sound sources and scenarios. The state vectors of the sound sources ranked first in the selection sequence or whose scores have changed significantly are input into the state machines for matching. When the HDMI state vector changes from ${Active: Playing}$ to ${Active: Silent, CEC: Standby}$, and the duration meets the condition, the state machine triggers a transition, generating an event signal labeled I_tv_sleep (TV sleep). When WiFi detects stream establishment and the buffer grows steadily, an I_cast_start (cast start) label is generated. The raw data changes are abstracted into high-level user intent events. This allows the system to understand the meaning behind user behavior (e.g., "turning off the TV means changing the song"), thereby achieving predictive switching preparation and greatly improving response speed.

[0038] The system updates denoising rules based on a preset state machine, denoising new sound source signals with event trigger tags to generate sound source confirmation result information. Debounce: A time window check is performed on tagged events. For example, the Bluetooth A2DP_START event must last for more than 2 seconds (the debounce window) to be considered valid; otherwise, it is judged as interference. Hysteresis sets a switching threshold $H$, and confirmation is only granted when the score of the new sound source is significantly higher than that of the current sound source ($Score_{new}>Score_{curr} + H$), preventing the signal from repeatedly jumping around the critical value. The number of switching times in a short period is counted; if it is too frequent, the switching logic is temporarily locked. Events that have passed the above checks are finally converted into "sound source confirmation result information," improving the robustness and stability of the system. It effectively filters noise caused by signal fluctuations, transient interference, and misoperation, ensuring the smoothness of "seamless switching" and avoiding user dizziness caused by repeated jumping between sound sources. Before sending the command, the target sound source's decoder (DSP), synchronization clock, and audio buffer are pre-initialized to ensure that cold start latency is minimized. A seamless transition is initiated by sending a switching command to the audio processing module. If the sampling rates are consistent, a direct crossfade-in / fade-out transition occurs; otherwise, resampling or time stretching is performed before mixing and output. A successful connection status is then fed back to update the global state matrix.

[0039] The above method establishes a multi-source parallel monitoring model and a global state matrix to achieve dynamic weighted scoring decisions based on source activity, content type semantics, and signal quality. It effectively suppresses transient signal interference and frequent jumps by using hysteresis control rules and double-layer labeling denoising technology. Combined with user intent prediction and decoding resource preloading mechanism, it eliminates the cold start delay of the audio path to achieve seamless transition. Furthermore, by adjusting model parameters and user preference weights through real-time feedback, the system has the ability to adaptively optimize for complex acoustic environments and personalized usage habits.

[0040] In a more specific embodiment, the execution method S110 further includes execution steps S111-S112.

[0041] S111, Sensing available signals from intelligent audio devices to obtain access information for sound source devices.

[0042] S112. Determine whether the status parameters in the sound source device access information are preset status values ​​to determine whether the smart audio device is in a pre-access state.

[0043] Specifically, multimodal signal scanning: The underlying firmware or driver layer of the smart audio device simultaneously activates multiple sensor modules and protocol stacks to perform real-time scanning of the entire frequency band and physical interfaces. Specific sensing methods include: Physical link sensing: For HDMI interfaces, monitoring the level changes of the HPD (Hot Plug Detect) pin and reading EDID (Extended Display Identifier) ​​data through the DDC channel. Wireless protocol sensing: For WiFi, listening to SSDP (Simple Service Discovery Protocol) broadcast messages, mDNS (Multicast DNS) service announcements, or handshake packets on specific ports of AirPlay / Chromecast within the local area network; for Bluetooth, enabling scanning mode and listening for ACL (Asynchronous Connection-Less) connection requests or pairing requests from surrounding devices. Monitoring the mounting status of internal I2S, PCM buses, or USB audio buses. The sensed raw signals (such as level transitions and data packet payloads) are parsed into structured "sound source device access information," which includes at least: device ID, device type (TV / Phone / PC), physical interface identifier, and connection timestamp. It achieves wide-area perception capability for multi-source heterogeneous devices. Through parallel signal sensing, devices no longer passively wait for a single command, but can actively discover any new devices attempting to connect in the environment, ensuring the real-time and comprehensiveness of sound source discovery and laying the physical foundation for subsequent rapid access. S112: Determine whether the status parameters in the sound source device access information are preset status values ​​to determine whether the intelligent audio device is in a pre-access state. Parameter extraction and comparison: The system parses the access information obtained in S111, extracts key status parameters, and compares them with the "preset status value" table stored in the system.

[0044] The specific judgment logic is as follows: If HPD_Level == High and CEC_Logical_Address is valid (not 0.0.0.0), it is determined to meet the preset status value and is marked as "pre-access". If Protocol_Handshake_Status == Completed but Audio_Stream_Rate == 0 (no audio data stream yet), it is determined to meet the preset status value and is marked as "pre-access". If ACL_Link_Status == Connected but A2DP_State != Streaming (not in streaming media transmission state), it is determined to meet the preset status value and is marked as "pre-access". Pre-access specifically refers to a device that has established a connection at the physical or link layer, completed identity verification, and has the ability to transmit audio at any time, but may currently be in a silent, paused, or waiting state for playback instructions. By comparing the preset status values, the system can filter out interference signals that are merely unpaired devices within range or have only made contact but not completed the handshake, and only retain devices that are truly "online and ready" for subsequent monitoring. By including "pre-connected" devices in the monitoring scope, the system can prepare for switching the device the instant before it actually starts playing (such as within milliseconds when the user presses the play button on their phone), achieving zero-latency tracking from the "connected" to the "playing" state.

[0045] In a more specific embodiment, before executing method S120, steps S121-S122 are also specifically included.

[0046] S121. Store the sound source category information of the intelligent audio device in the database to create a multi-sound source parallel state monitoring model.

[0047] S122. Based on the degree of matching between the sound source category information of the intelligent audio device and the user's habit and preference needs, arrange the sound source selection agenda for the intelligent audio device in the multi-sound source parallel state monitoring model.

[0048] Specifically, the initialization and pre-configuration preceding step S120 proceed. In step S121, the sound source category information of the smart audio device is stored in the database to create a multi-source parallel state monitoring model. The sound source metadata structure is constructed as follows: During the initialization phase, the system first defines metadata templates for all supported sound source categories. This information includes: Sound source type identifiers: such as S_HDMI, S_WIFI_AirPlay, S_BT_A2DP, etc. Physical / protocol characteristics: port number, protocol stack type (TCP / UDP / L2CAP), supported audio encoding formats (PCM / AC3 / AAC / LDAC), heartbeat keep-alive interval, timeout threshold, buffer size, etc. Database storage and instantiation: This structured sound source category information is written to the configuration database in the device's non-volatile memory (such as Flash or EEPROM). Upon system startup, the above information is read from the database, and the "multi-source parallel state monitoring model" is instantiated in memory. The model is essentially a state container with several empty slots, each corresponding to a sound source category, and it reserves a data structure interface for storing the real-time state vector $V_i(t)$. Through database-driven model creation, sound source management is standardized. Adding a new sound source type (such as adding SpotifyConnect) only requires updating the database configuration, without modifying the underlying monitoring code, thus improving system scalability. The monitoring framework is determined at the beginning of device operation, providing stable, pre-allocated memory space and logical anchors for subsequent parallel access and data acquisition of multiple sound sources, avoiding dynamic resource allocation conflicts during runtime. S122: Based on the degree of matching between the sound source category information of the smart audio device and user habit preferences, the smart audio devices in the multi-source parallel state monitoring model are arranged according to a sound source selection agenda. User habit preference reading: The system reads the user's historical usage data and preference configuration from local storage or cloud account. Preference data includes historical priority corrections, such as users habitually prioritizing mobile WiFi music over TV during "evening" hours. Manual intervention records the frequency and direction of past manual forced switching by the user. Scene tags such as "movie mode," "music mode," and "night mode" are used. The system matches "sound source category information" from the database with "user habit preferences." If the current time period is "evening," and the sound source category is "WiFi streaming media," and "evening WiFi streaming media" is frequently used by users, a high match is calculated. The optimal agenda is pre-arranged, dynamically adjusting the "basic priority" or agenda rules to be loaded in step S120 based on the match calculation results. The system generates an initial "optimal agenda sequence," for example, temporarily elevating a WiFi sound source that was originally ranked second by default to first place during a specific time period. The decision logic is "pre-set" based on user habits before the actual sound source signal is received (before S110 / S120 execution).This allows the system to be in a "user-aware" state from the moment it powers on, rather than relying solely on real-time calculations. By pre-arranging the agenda based on preferences, the computational workload of complex weight adjustments during the subsequent real-time monitoring phase (S150) is reduced, resulting in a more agile sound source switching response. This ensures that the sound source switching logic is highly consistent with the user's psychological expectations, minimizing situations where the system's automatic judgment contradicts the user's subjective intent.

[0049] In a more specific embodiment, sound source status judgment information of each sound source signal is obtained, and a sound source status vector is established based on the sound source status judgment information, including performing vector feature processing on the sound source status judgment information to obtain the sound source status vector; judging whether the sound source status vector conforms to the preset sound source demand matching rule, so as to determine whether the sound source signal corresponding to the sound source status vector is in the access state.

[0050] Before selecting a connectable state sound source signal from the global state matrix of the sound source to add to the sound source selection sequence, the method also includes prioritizing the sound source signals based on the degree of matching between the connectable state sound source signals and the preset audio stability requirement information.

[0051] Specifically, the sound source state judgment information is processed into a vector feature to obtain the sound source state vector. Feature extraction and quantization: The raw state judgment information obtained by the system from the underlying driver and protocol stack is often heterogeneous (such as CEC bytes, socket state, RSSI integer values). The system standardizes and quantizes this information, mapping it into continuous or discrete feature values. For example: RSSI (signal strength) is mapped to a signal quality feature $Q$ in the range $[0, 1]$; audio packet arrival rate is mapped to an activity feature $A$; SDP or EDID information is parsed to extract content type features $C$ (such as Music=0, Video=1); Extract timestamp features $T$. Combine the quantized features into a multi-dimensional vector $V_i$ according to predefined dimensions. For example, construct a five-dimensional vector: $V_i = [LinkState, ActiveState, ContentType, SignalQuality, TimeDelta]$. The states of all sound sources are converted into vectors of this unified format to facilitate matrix operations and machine learning model processing. Data standardization unifies data from different protocols and physical layers into mathematical objects (vectors) that are easy for computers to process, eliminating processing barriers caused by heterogeneous data. Digital representation makes the originally ambiguous "state" measurable and computable, laying the data foundation for subsequent accurate scoring and matching. Determine whether the sound source state vector conforms to the preset sound source demand matching rules to determine whether the sound source signal corresponding to the sound source state vector is in an accessed state. The system presets a set of "sound source demand matching rules", which are the logical thresholds for determining whether a sound source is truly "effectively accessed". This is not just a physical connection, but refers to the ability to continuously play audio.

[0052] Specific determination logic - Link integrity check: If $LinkState \neq Connected$, it is determined that the rule is not met. Service availability check: If $ActiveState = Silent$ and the duration exceeds the silent threshold (e.g., 5 seconds), or $SignalQuality<Th_{min}$ (e.g., Bluetooth RSSI < -85 dBm), it is determined that this vector does not meet the requirements of an "effective sound source". For a WiFi sound source, it is determined to meet the rule only when the RTSP / HTTP SETUP interaction is completed and the Session ID is valid. Only when all key dimensions of the vector $V_i$ pass the matching rule verification, the system marks this sound source as "connected state" and allows it to enter the global state matrix to participate in subsequent competition; otherwise, it is marked as "unavailable" or "standby", and is isolated or ignored. "False connections" or "zombie connections" (such as Bluetooth headsets that are paired but too far away, or devices that have established a WiFi connection but have not been screen-cast) are effectively eliminated, avoiding interference from invalid sound sources to system decisions. Ensure that only sound sources with real playback capabilities are included in the scheduling scope, preventing system misjudgment caused by the state jitter of invalid sound sources. According to the matching degree between the sound source signal in the connectable state and the preset audio stability requirement information, the audio stability priority of the sound source signal is arranged. Extract stability indicators. For sound sources determined to be in the "connectable state", the system further extracts the stability indicators of their physical layer and link layer. Including: WiFi: packet loss rate, jitter value, bandwidth margin; Bluetooth: RSSI attenuation rate, connection interval stability, number of retransmissions; HDMI: clock jitter, error frame count. These indicators are compared with the preset "audio stability requirement information" (i.e., the minimum quality threshold for various coding formats required for the normal operation of the system). For example, when playing high-bitrate lossless music, the stability requirement is that the WiFi bandwidth > 2 Mbps and the packet loss rate < 0.01%. If the indicators of a certain sound source do not meet the requirements, its "matching degree" is marked as low. According to the level of the matching degree, the candidate sound sources are pre-sorted. Sound sources with extremely high signal quality (matching degree 100%) are ranked at the front; sound sources with barely meeting signal quality (matching degree 60% - 90%) are ranked in the middle and marked with a "high risk" label; sound sources with signal quality not meeting the requirements (matching degree < 60%), even if they are connected, are removed from the candidate list of the "sound source selection sequence" or placed at the lowest priority "guaranteed position". Before deciding "who to switch to", first ensure that "it can be listened to well after switching". Prevent the system from switching to a sound source with unstable signal, which will cause stuttering or interruption, just because of high priority. Avoid the bad experience of "stuttering after switching", and ensure that the sound source after automatic switching is always the smoothest and most stable in the current environment.By conducting stability screening upfront, the number of low-quality candidates that need to be processed in the subsequent main optimization agenda (S150) is reduced, thereby improving the system's operating efficiency and response speed.

[0053] According to the preset state machine update denoising rules, new sound source signals with event trigger tags are denoised to generate sound source confirmation result information. This includes assigning user habit fit priority values ​​to the sound source category information of the sound source signals in the sound source selection sequence; arranging the priority according to the user habit fit priority values ​​to obtain the priority confirmation signal sequence corresponding to the sound source selection sequence; and performing event trigger tagging preprocessing on the sound source signals in the priority confirmation signal sequence.

[0054] The event-triggered labeling preprocessing is performed on the sound source signals in the priority confirmation signal sequence, including obtaining the sound source category information of at least one sound source signal in the priority confirmation signal sequence for initial target sound source labeling; obtaining the sound source category information of at least one sound source signal that has undergone initial target sound source labeling for advanced target sound source labeling; and generating sound source confirmation result information based on the sound source category information of the sound source signal that has undergone advanced target sound source labeling.

[0055] Specifically, the system extracts the current time (e.g., weekday evening), scenario (e.g., "watching a movie"), and category information (e.g., WiFi streaming, Bluetooth headphones) of each candidate sound source in the sound source optimization sequence. It retrieves historical switching data for this scenario from local storage or cloud-based user profiles. The system calculates the matching degree between the current sound source and the user's habits. For example, if historical user data shows an 80% probability of switching from HDMI to WiFi music during "20:00-22:00", the WiFi sound source is assigned a high matching weight; if the user rarely answers Bluetooth calls while watching TV, the Bluetooth sound source is assigned a low matching weight. The calculated matching degree is converted into a specific priority correction value and added to the base score of the sound source signal. Based on the priority assignment value according to the user's habit matching degree, a priority confirmation signal sequence is obtained corresponding to the sound source optimization sequence. The sequence is reordered according to the priority correction value obtained in step 1, re-sorting the original sound source optimization sequence. A new sequence is generated and output as a new "priority confirmation signal sequence". This sequence considers not only the physical state of the signal (activity, quality) but also the user's personalized preferences. Sound sources with high compatibility are moved to the front of the sequence, while those with low compatibility are moved to the back or removed. This breaks away from the traditional "one-size-fits-all" switching logic of devices, achieving intelligent decision-making based on individual user habits, greatly improving the accuracy of automatic switching and user satisfaction. Pre-decision optimization filters out sound sources that are unlikely to conform to the user's intentions before performing complex signal denoising processing, reducing the computational load on subsequent processing units.

[0056] Event-triggered labeling preprocessing is performed on the sound source signals in the priority confirmation signal sequence. For the leading candidate sound sources in the "priority confirmation signal sequence," a two-stage labeling process is initiated. This provides high-dimensional feature labels for subsequent denoising decisions, enabling the system to identify the semantic meaning of the signal (whether it's music or noise) and behavioral patterns (whether it's a mis-touch or an operation). Transforming cold, impersonal electrical signals into "labels" with business meaning provides accurate judgment criteria for denoising rules, avoiding misjudgments caused by relying solely on signal strength.

[0057] Obtain the sound source category information of the sound source signal in at least one priority confirmation signal sequence for initial target sound source labeling processing. Coarse-grained classification: Perform the first round of labeling on the sound source based on protocol layer information and basic metadata.

[0058] Tag definition: Identified as [call type] (such as Bluetooth SCO / HFP protocol); Identified as [media playback type] (such as A2DP / AirPlay audio stream); Identified as [system prompt tone type] (such as short prompt tone, navigation voice); It is identified as [video / audio track] (such as HDMI / eARC audio).

[0059] For [call-related] tags, the highest priority pass flag is directly assigned; for [system prompt tone-related] tags, a low priority flag is directly assigned. By quickly distinguishing between "important events" (such as phone calls) and "interference events" (such as notifications) through simple protocol characteristics, noise reduction efficiency is greatly improved, preventing critical communications from being missed.

[0060] The system acquires the sound source category information of at least one sound source signal that has undergone initial target sound source tagging and then performs advanced target sound source tagging. Building upon the initial tags, it combines temporal features and state machine transitions for deeper tagging. For signals labeled "[Media Playback]", its duration is calculated. If the duration is less than 2 seconds, it is tagged with "[Instantaneous Noise]". For HDMI signals, if CEC Standby status is detected, it is tagged with "[About to be Released]". For WiFi signals, if the buffer fluctuates drastically, it is tagged with "[Unstable Stream]", and the historical window is used to determine whether the signal is intermittent interference (such as Bluetooth momentary disconnection and reconnection). Combined tags are generated, such as "[Media Playback] + [Stable Stream]", or "[Notification Tone] + [Instantaneous Noise]". This effectively solves the problem that traditional technologies cannot distinguish between "brief false triggers" and "true playback intent". Through advanced tagging, the system can accurately identify and filter out interference signals such as Bluetooth ringtones and mobile phone notification sounds, ensuring that only genuine and stable playback intent is retained.

[0061] Based on the sound source category information of the sound source signal after advanced target sound source tagging processing, a sound source confirmation result is generated. If the tag is [Call Type] or [Media Playback Type] + [Stable Stream], and it ranks high in the priority confirmation signal sequence, a [Confirm Switching] result is generated. If the tag is [Temporary Noise] or [System Prompt Tone Type] or [Unstable Stream], a [Reject Switching / Maintain Current] result is generated. If the tag is [About to be Released], a [Waiting for Takeover] result is generated, and a check of the secondary priority sound source is triggered. This "sound source confirmation result" serves as the final instruction, passed to the execution module (S180) for physical switching of the audio path. With the dual guarantee of "habit fit sorting" and "double-layer tagging filtering," the generated confirmation result has extremely high accuracy. It truly achieves "seamless" switching—a millisecond-level response when switching is appropriate (e.g., phone call), and no movement when switching is inappropriate (e.g., accidental notification touch), perfectly solving the problem of fragmented interaction under multiple sound source coexistence.

[0062] Before arranging the sound source signals in the sound source selection agenda in the sound source selection sequence, the method further includes pre-labeling the sound source signals with the signal transmission medium type to obtain a pre-labeled sound source signal sequence; and summarizing the adaptation priority of each sound source signal according to the pre-labeled sound source signal sequence to obtain the sound source selection sequence.

[0063] After performing state machine matching on the sound source signals in the sound source selection agenda to obtain new sound source signals with event trigger labels, the method includes performing predictive filtering processing on the sound source signals according to preset user intent filtering rules. The types of sound source signals after predictive filtering processing include original intent sound source signals, optimized switching intent sound source signals, and user potential intent sound source signals. The original intent sound source signals, optimized switching intent sound source signals, and user potential intent sound source signals are compared internally and ranked by intent degree, excluding the sound source signals with the lowest user intent degree; the event trigger labels are updated on the remaining sound source signals.

[0064] Before denoising a new sound source signal with an event-triggered tag to generate sound source confirmation result information according to the preset state machine update denoising rules, the method includes: configuring preset hysteresis control rules in a multi-sound-source parallel state monitoring model; calculating the connection time of the sound source signal according to the hysteresis control rules, wherein the hysteresis control rules include a hysteresis threshold, the hysteresis threshold is used to perform pre-timing tagging processing on the sound source signal, and retaining sound source signals that simultaneously meet the hysteresis control rules and the updated denoising rules.

[0065] After denoising new sound source signals with event trigger tags according to the preset state machine update denoising rules to generate sound source confirmation result information, the method includes judging the sound source activity of the sound source signals in the sound source confirmation result information according to the preset sound source activity judgment rules to obtain sound source activity judgment information; obtaining the activity score, content type score and signal quality score of the sound source signal according to the sound source activity judgment information and configuring them in the multi-sound source parallel state monitoring model for configuration model training.

[0066] Specifically, the preset hysteresis control rules are configured in the multi-source parallel state monitoring model.

[0067] Rule injection: Before or during system operation, the logic module containing the hysteresis algorithm is loaded into the kernel of the "multi-source parallel state monitoring model".

[0068] Parameter definition: Configure specific hysteresis parameters, including: Switch confirmation time ($T_{confirm}$): The shortest time (e.g., 1.5 seconds) during which the sound source must remain stable.

[0069] Hysteresis threshold ($H$): The score of the new sound source must exceed the difference between the score of the current sound source and the score of the current sound source (e.g., if the current score is 80, the new sound source must be >85 to trigger).

[0070] Lock time ($T_{lock}$): The time window during which switching is prohibited after switching.

[0071] Technical effects: Build a defense mechanism: Install an "anti-vibration guardrail" on the monitoring model to ensure that the subsequent denoising steps are carried out in a stable and margin-based discrimination environment, thus avoiding the uncertainty of the critical point from the bottom logic level.

[0072] The duration of the connection for the sound source signal is calculated based on hysteresis control rules. These rules include a hysteresis threshold, which is used for pre-timing tagging of the sound source signal. Continuous timing: When the monitoring model detects that a sound source requests to switch (such as Bluetooth A2DP starting streaming), it starts an internal timer to calculate the duration for which the sound source remains in a "connected and active" state.

[0073] Label processing before timing: Pre-labeling: Before the timer reaches $T_{confirm}$, the system assigns a temporary "timer in progress" or "pending" label to the sound source.

[0074] Hysteresis comparison: Calculate $Score_{new} - Score_{current}$ in real time. If the difference does not reach the hysteresis threshold $H$, a "failed" label is added even if the timer is completed.

[0075] Status freeze: Before the "Timer" label is removed, even if the sound source score is temporarily high, the system will not trigger the final confirmation of the noise reduction process and will force it to enter the waiting state.

[0076] Based on preset sound source activity judgment rules, the sound source signal within the sound source confirmation result information is judged for sound source activity to obtain sound source activity judgment information. After the system issues sound source confirmation result information (i.e., decides to switch or hold), the process does not immediately end, but continues to track the actual performance of the sound source. Using the "sound source activity judgment rules," the actual behavior of the sound source over a subsequent period is analyzed: Should the video continue playing (to verify "genuine activity")? Did it stop playing after 1 second (to verify "false trigger")? Does the game frequently freeze (to verify "poor quality")? Did the user manually switch back to the original audio source within 5 seconds (verify "User intent recognition error")? The above tracking results are encapsulated into "sound source activity judgment information," and the "success" or "failure" attribute of this switching decision is marked. Decision feedback: A "decision-verification" feedback loop is established, allowing the system to perceive whether each automatic switch was truly "correct."

[0077] Based on the sound source activity assessment information, the activity score, content type score, and signal quality score of the sound source signal are obtained and configured into the multi-sound source parallel state monitoring model for configuration model training. Feature extraction and quantization: Based on activity level information, the system extracts key feature values ​​of the sound source in reverse: If playback is stable, extract the high activity score ($A$); If a specific type is identified (such as VoIP), record the content type score ($C$); If the stuttering is severe, record the low signal quality score ($Q$). Use this score to fine-tune the weighting coefficients in the scoring function $Score = w_p P + w_a A + \dots$. If a user manually switches back to WiFi music due to slightly low signal quality, it indicates that $w_q$ (quality weight) may be too low and needs to be increased. Dynamically adjust the hysteresis threshold $H$ or the confirmation time $T_{confirm}$ based on historical data. Write the new parameters obtained after training into the configuration file of the "Multi-Source Parallel State Monitoring Model," overwriting the old parameters. The device can automatically optimize its judgment logic based on the actual acoustic environment of the home, network conditions, and the user's unique usage habits. Over time, the model's scoring mechanism will become increasingly closer to the user's preferences; for example, automatically lowering the hysteresis threshold for users with high tolerance (more sensitive), and automatically increasing the quality weight for users who demand high sound quality (more discerning), thus achieving a truly intelligent experience.

[0078] Figure 4 This is a schematic block diagram of a sound source switching device based on priority judgment provided in an embodiment of this application. As shown in the figure, corresponding to the above-described sound source switching method based on priority judgment, this application also provides a sound source switching device 100 based on priority judgment. This sound source switching device based on priority judgment includes a unit for executing the above-described sound source switching method based on priority judgment, and the device can be configured in a desktop computer, tablet computer, laptop computer, or other terminal. Specifically, please refer to... Figure 4The priority-based sound source switching device 100 includes a first judgment unit 110, used to acquire the sound source device access information of the smart audio device and determine whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information; a priority arrangement unit 120, used to select the sound source signals in the pre-access state and add them to a preset multi-sound source parallel state monitoring model, in which the smart audio device arranges the sound source connection priority; a second judgment unit 130, used to acquire the sound source state judgment information of each sound source signal and establish a sound source state vector based on the sound source state judgment information; and a state matrix generation unit 140, used to merge the sound source state vectors to generate a state matrix. The system comprises: a global state matrix for the sound source; an agenda arrangement unit 150, used to select connectable sound source signals from the global state matrix and add them to the sound source selection sequence, where the sound source signals are arranged in the sound source selection agenda; a tagging unit 160, used to perform state machine matching on the sound source signals in the sound source selection agenda to obtain new sound source signals with event trigger tags; a denoising unit 170, used to update the denoising rules according to the preset state machine and denoise the new sound source signals with event trigger tags to generate sound source confirmation result information; and an instruction generation unit 180, used to generate and send confirmation connection instructions for the sound source signals that match the smart audio device based on the sound source confirmation result information.

[0079] In addition, the device further includes an information acquisition unit for sensing available signals of the smart audio device to obtain sound source device access information; and a determination unit for determining whether the status parameters in the sound source device access information are preset status values, so as to determine whether the smart audio device is in a pre-access state.

[0080] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned priority-based sound source switching device and each unit can be referred to the corresponding description in the foregoing method embodiments. For the sake of convenience and brevity, it will not be repeated here.

[0081] The aforementioned priority-based sound source switching device can be implemented as a computer program, which can, for example, Figure 5 It runs on the computer device shown.

[0082] Please see Figure 5 This diagram illustrates a schematic block diagram of a computer device provided in an embodiment of this application. The computer device 500 can be a terminal or a server. The terminal can be an electronic device with communication functions, such as a smartphone, tablet, laptop, desktop computer, personal digital assistant, or wearable device. The server can be a standalone server or a server cluster consisting of multiple servers.

[0083] The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. The memory may include a non-volatile storage medium 503 and internal memory 504.

[0084] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions that, when executed, cause the processor 502 to perform a priority-based sound source switching method.

[0085] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0086] The internal memory 504 provides an environment for the execution of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a sound source switching method based on priority judgment.

[0087] This network interface 505 is used for network communication with other devices. Those skilled in the art will understand that... Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device 500 to which the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0088] It should be understood that in the embodiments of this application, the processor 502 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0089] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0090] Therefore, this application also provides a storage medium. This storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein the computer program includes program instructions. When executed by a processor, the program instructions cause the processor to perform the following steps: The system acquires the sound source device access information of the smart audio device and determines whether each sound source signal within the smart audio device is in a pre-access state based on this information. Sound source signals in the pre-access state are added to a preset multi-source parallel state monitoring model, where the smart audio device prioritizes sound source connections. The system acquires the sound source state judgment information for each sound source signal and establishes a sound source state vector based on this information. The sound source state vectors are merged to generate a global sound source state matrix. Sound source signals in the global sound source state matrix that are in a connectable state are added to a sound source optimization sequence, where they are arranged in a sound source optimization agenda. State machine matching is performed on the sound source signals in the sound source optimization agenda to obtain new sound source signals with event trigger tags. The denoising rules are updated according to the preset state machine, and the new sound source signals with event trigger tags are denoised to generate sound source confirmation result information. Based on the sound source confirmation result information, a confirmation connection command is generated and sent to the sound source signal that matches the smart audio device.

[0091] The storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0092] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0093] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For example, the division of each unit is merely a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0094] The steps in the methods of this application embodiment can be adjusted, merged, or deleted according to actual needs. The units in the apparatus of this application embodiment can be merged, divided, or deleted according to actual needs. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0095] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a 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 includes several instructions to cause a computer device (which may be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0096] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for sound source switching based on priority judgment, applied to a smart audio device, comprising the steps of: The method includes: Obtain the sound source device access information of the smart audio device, and determine whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information; The sound source signal in the pre-access state is selected and added to the preset multi-source parallel state monitoring model. The intelligent audio device arranges the sound source connection priority in the multi-source parallel state monitoring model. Acquire sound source state judgment information for each sound source signal, and establish a sound source state vector based on the sound source state judgment information; The state vectors of each sound source are merged to generate a global state matrix of the sound sources; The sound source signal with a connectable state is selected from the global state matrix of the sound source and added to the sound source selection sequence. The sound source signal is then arranged in the sound source selection sequence according to the sound source selection agenda. The sound source signals in the sound source selection agenda are matched using a state machine to obtain new sound source signals with event trigger tags; According to the preset state machine update denoising rules, the new sound source signal with the event trigger tag is denoised to generate sound source confirmation result information; Based on the sound source confirmation result information, a confirmation connection command is generated and sent for the sound source signal that matches the smart audio device.

2. The priority judgment based sound source switching method according to claim 1, wherein, The step of obtaining the sound source device access information of the smart audio device, and determining whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information, includes: The intelligent audio device is subjected to available signal sensing to obtain sound source device access information; Determine whether the status parameter in the access information of the sound source device is a preset status value to determine whether the smart audio device is in a pre-access state.

3. The priority-based sound source switching method according to claim 2, wherein, Before the sound source signal in the selected pre-access state is added to the preset multi-source parallel state monitoring model, the method includes: The sound source category information of the intelligent audio device is stored in the database to create a multi-sound source parallel state monitoring model; Based on the degree of matching between the sound source category information of the intelligent audio device and the user's habitual preference needs, the intelligent audio device in the multi-sound source parallel state monitoring model is arranged according to the sound source optimization agenda.

4. The priority-based sound source switching method according to claim 3, wherein, The step of acquiring sound source state judgment information for each sound source signal and establishing a sound source state vector based on the sound source state judgment information includes: The sound source state judgment information is processed into a vector feature to obtain a sound source state vector; Determine whether the sound source state vector conforms to the preset sound source demand matching rules, so as to determine whether the sound source signal corresponding to the sound source state vector is in an accessed state.

5. The priority-based sound source switching method according to claim 4, wherein, Before adding the sound source signal with a connectable state selected in the global state matrix of the sound source to the sound source preference sequence, the method further includes: Based on the degree of matching between the sound source signal in the connectable state and the preset audio stability requirement information, the sound source signal is prioritized according to audio stability.

6. The priority judgment based sound source switching method according to claim 5, wherein, The step of updating the denoising rules according to a preset state machine and denoising the new sound source signal with the event trigger tag to generate sound source confirmation result information includes: The sound source category information of the sound source signals in the sound source selection sequence is assigned a priority value based on the degree of user habit fit. The priority sequence is obtained by assigning priority values ​​to users based on the degree of fit of their habits, and then arranging them according to the priority sequence of the sound source selection. The sound source signals in the priority confirmation signal sequence are preprocessed with event-triggered tagging.

7. The priority judgment based sound source switching method according to claim 6, wherein, The event-triggered tagging preprocessing of the sound source signals in the priority confirmation signal sequence includes: Obtain sound source category information of at least one of the sound source signals in the priority confirmation signal sequence and perform initial target sound source labeling processing; Acquire the sound source category information of at least one sound source signal that has undergone the initial target sound source labeling process, and then perform advanced target sound source labeling process; Based on the sound source category information of the sound source signal that has undergone the advanced target sound source tagging process, a sound source confirmation result information is generated.

8. The priority judgment based sound source switching method according to claim 1, wherein, Before the sound source signal is arranged in the sound source selection agenda in the sound source selection sequence, the method further includes: The sound source signal is pre-labeled with the signal transmission medium type to obtain a pre-labeled sound source signal sequence; The sound source optimization sequence is obtained by summarizing the adaptation priority of each sound source signal according to the pre-labeled sequence of the sound source signals.

9. The priority judgment based sound source switching method according to claim 1, wherein, After performing state machine matching on the sound source signals in the sound source selection agenda arrangement to obtain new sound source signals with event trigger labels, the method includes: The sound source signal is predicted and filtered according to the preset user intent filtering rules. The types of the sound source signal after the prediction and filtering process include the original intent sound source signal, the optimized switching intent sound source signal, and the user's potential intention sound source signal. The original intention sound source signal, the optimized switching intention sound source signal, and the user's potential intention sound source signal are compared internally and ranked according to their degree of intent. The sound source signal with the lowest degree of user intent is then excluded. The remaining sound source signals are updated with event-triggered tags.

10. The priority judgment based sound source switching method according to claim 1, wherein, Before updating the denoising rules according to a preset state machine and denoising the new sound source signal with the event trigger tag to generate sound source confirmation result information, the method includes: The preset hysteresis control rules are configured in the multi-source parallel state monitoring model; The duration of the connection of the sound source signal is calculated according to the hysteresis control rule, which includes a hysteresis threshold. The hysteresis threshold is used to perform pre-timing labeling on the sound source signal and retain the sound source signal that simultaneously meets the hysteresis control rule and the updated denoising rule.

11. The priority judgment based sound source switching method according to claim 1, wherein, After updating the denoising rules according to a preset state machine and denoising the new sound source signal with the event trigger tag to generate sound source confirmation result information, the method includes: According to the preset sound source activity judgment rules, the sound source signal in the sound source confirmation result information is judged to determine the sound source activity, so as to obtain the sound source activity judgment information. Based on the sound source activity judgment information, the activity score, content type score, and signal quality score of the sound source signal are obtained and configured in the multi-sound source parallel state monitoring model for configuration model training.

12. A priority-based sound source switching device, characterized by comprising: The device includes: The first judgment unit is used to obtain the sound source device access information of the smart audio device and determine whether each sound source signal in the smart audio device is in a pre-access state based on the sound source device access information. A priority sorting unit is used to select the sound source signal in the pre-access state to be added to a preset multi-sound source parallel state monitoring model, and the intelligent audio device sorts the sound source connection priority in the multi-sound source parallel state monitoring model. The second judgment unit is used to obtain the sound source state judgment information of each sound source signal and to establish a sound source state vector based on the sound source state judgment information. The state matrix generation unit is used to merge the state vectors of each sound source to generate a global state matrix of the sound sources; An agenda arrangement unit is used to select the sound source signal with a connectable state from the global state matrix of the sound source and add it to the sound source selection sequence, wherein the sound source signal is arranged in the sound source selection agenda in the sound source selection sequence. A tagging unit is used to perform state machine matching on the sound source signals in the sound source selection agenda arrangement to obtain new sound source signals with event triggering tags; The denoising unit is used to update the denoising rules according to the preset state machine, and to denoise the new sound source signal with the event trigger tag to generate sound source confirmation result information. The instruction generation unit is used to generate and send a confirmation connection instruction for the sound source signal that matches the smart audio device based on the sound source confirmation result information.

13. The priority judgment based sound source switching apparatus according to claim 12, wherein, The device further includes: An information acquisition unit is used to sense available signals from the intelligent audio device to obtain sound source device access information; The determination unit is used to determine whether the status parameter in the access information of the sound source device is a preset status value, so as to determine whether the smart audio device is in a pre-access state.

14. A computer device, comprising: The computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method as described in any one of claims 1-11.

15. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which includes program instructions that, when executed by a processor, can implement the method as described in any one of claims 1-11.

Images