Conversation state recovery method and device, wearable equipment and storage medium

Abstract

The invention provides a dialogue state recovery method and device, wearable equipment and a storage medium, and belongs to the technical field of intelligent communication, and the method comprises the steps: determining a first connection state of a first information transmission channel between the wearable equipment and terminal equipment; obtaining a dialogue interruption data packet in response to the fact that the first connection state is a disconnected state; detecting a second connection state of a second information transmission channel between the wearable device and the terminal device in response to the condition that the first connection state is not recovered from the disconnected state to the connected state within a third preset time period; and switching an information transmission channel between the wearable device and the terminal device from the first information transmission channel to a second information transmission channel in response to the connection state of the second connection state, and receiving reply information fed back by the terminal device through the second information transmission channel. According to the dialogue state recovery method and device, the wearable device and the storage medium, the dialogue state can be reliably recovered when the connection is interrupted.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent communication, and more particularly relates to a dialogue state recovery method and device, a wearable device, and a storage medium. BACKGROUND

[0002] With the popularization of Internet of Things technology, wearable devices have become an important portal for people to interact with the digital world. Among them, the voice dialogue function is favored due to its convenience. Usually, these devices serve as a front-end interaction interface, working in cooperation with terminal devices such as smart phones through Bluetooth connection, and the terminal devices carry powerful natural language processing models to complete dialogue understanding and reply generation.

[0003] However, this architecture is highly dependent on the stability of Bluetooth connection. In actual application scenarios, Bluetooth connection is prone to accidental interruption due to user movement, environmental interference, or device power consumption management, etc. When handling such interruptions, existing technologies often adopt a simple reconnection mechanism. If reconnection fails, the dialogue service is forced to terminate, resulting in a sharp decline in user experience. The more prominent problem is that the user may be inputting voice instructions or the device may be playing back reply information returned by the terminal when the interruption occurs. This "silent" interruption processing method will directly cause the loss of dialogue context and the rupture of semantic coherence. SUMMARY

[0004] The purpose of the present application is to provide a dialogue state recovery method and device, a wearable device, and a storage medium to reliably recover the dialogue state when the connection is interrupted.

[0005] The first aspect of the embodiment of the present application provides a dialogue state recovery method applied to a wearable device, comprising: In the process of dialogue between the wearable device and the terminal device, the first connection state of the first information transmission channel between the wearable device and the terminal device is determined; the first information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Bluetooth connection; In response to the first connection state being in a disconnected state, dialogue interruption data packets are obtained; the dialogue interruption data packets include historical dialogue information within a first preset time period before the disconnected state occurs, voice information of the user within a second preset time period after the disconnected state occurs, and respective corresponding time stamps; In response to the first connection state not recovering from the disconnected state to the connected state within a third preset time period, the second connection state of the second information transmission channel between the wearable device and the terminal device is detected; the second information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Wi-Fi connection, and the communication priority of the first information transmission channel is higher than that of the second information transmission channel; In response to the second connection state being in a connected state, the information transmission channel between the wearable device and the terminal device is switched from the first information transmission channel to the second information transmission channel, and the dialogue interruption data packet is sent to the terminal device through the second information transmission channel so that the terminal device generates corresponding response information based on the dialogue interruption data packet; The device receives a response from the terminal device via the second information transmission channel to restore the dialogue state between the wearable device and the terminal device.

[0006] A second aspect of this application provides a dialogue state recovery device, which is disposed in a wearable device, and the device includes: The first connection status determination module is used to determine the first connection status of the first information transmission channel between the wearable device and the terminal device during the dialogue between the wearable device and the terminal device; the first information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Bluetooth connection. The dialogue interruption data acquisition module is used to acquire dialogue interruption data packets in response to the first connection state being disconnected. The dialogue interruption data packets include historical dialogue information within a first preset time period before the disconnection occurred, the user's voice information within a second preset time period after the disconnection occurred, and their respective timestamps. The second connection status determination module is used to detect the second connection status of the second information transmission channel between the wearable device and the terminal device in response to the first connection status failing to recover from the disconnected state to the connected state within a third preset time period; the second information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Wi-Fi connection, and the communication priority of the first information transmission channel is higher than the communication priority of the second information transmission channel. The transmission channel switching module is used to switch the information transmission channel between the wearable device and the terminal device from the first information transmission channel to the second information transmission channel in response to the second connection state being in the connection state, and send the dialogue interruption data packet to the terminal device through the second information transmission channel so that the terminal device can generate corresponding reply information based on the dialogue interruption data packet; The dialogue state recovery module is used to receive the reply information from the terminal device through the second information transmission channel in order to restore the dialogue state between the wearable device and the terminal device.

[0007] A third aspect of this application provides a wearable device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described dialogue state restoration method.

[0008] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described dialogue state restoration method.

[0009] The beneficial effects of the dialogue state recovery method and apparatus, wearable device, and storage medium provided in this application are as follows: By creating and sending a dialogue interruption data packet integrating key contextual information before and after the disconnection, this application effectively solves the problem of dialogue content fragmentation that may occur when an interruption occurs, ensuring semantic continuity. Simultaneously, by introducing a "wait-detect-switch" intelligent channel switching mechanism, and prioritizing low-power Bluetooth, unnecessary switching caused by momentary interference is avoided, ensuring device energy efficiency. Furthermore, when Bluetooth cannot be restored, Wi-Fi can be seamlessly activated as a backup channel, ensuring continuous communication capabilities. Ultimately, this overcomes the drawbacks of silent interruption handling in existing technologies, enabling intelligent, coherent, and reliable recovery of dialogue services between wearable devices and terminal devices after unexpected connection interruptions, greatly improving the user's interactive experience and service reliability in mobile scenarios. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, 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 some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A flowchart illustrating a dialogue state restoration method provided in an embodiment of this application; Figure 2 This is a structural block diagram of a dialogue state recovery device provided in an embodiment of this application; Figure 3 This is a schematic block diagram of a wearable device provided in an embodiment of this application. Detailed Implementation

[0012] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0013] It is understood that in the embodiments of this application, data such as user information are involved. When the embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with relevant laws, regulations and standards.

[0014] It should be noted that the terms "first," "second," etc., used in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein.

[0015] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0016] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a dialogue state restoration method provided in an embodiment of this application. The dialogue state restoration method provided in this embodiment can be executed by a wearable device, and the method may include: S101: During the dialogue between the wearable device and the terminal device, the first connection state of the first information transmission channel between the wearable device and the terminal device is determined; the first information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Bluetooth connection.

[0017] In this embodiment, the wearable device is an intelligent electronic device that can be directly worn on the human body, and has functions such as voice acquisition, audio playback, and data transmission. It can be smart glasses, etc.

[0018] The terminal device is the main device used in conjunction with the wearable device. It possesses enhanced computing, storage, and networking capabilities to process dialogue requests from the wearable device and generate responses. This device can be a smartphone, tablet, or similar device. The first information transmission channel is a wireless communication channel established between the wearable device and the terminal device based on the Bluetooth protocol. It is used to transmit voice, commands, and other dialogue data, featuring low power consumption and stable short-range communication, making it the preferred communication method during dialogue. The first connection state refers to the connection status of the first information transmission channel, including a connected state (i.e., the Bluetooth link is normal and data can be transmitted) and a disconnected state (i.e., the Bluetooth link is interrupted and data cannot be transmitted).

[0019] In this embodiment, during real-time dialogues such as voice conversations and command interactions between the wearable device and the terminal device, the wearable device can continuously monitor the connection status of the first information transmission channel between them. Since the first information transmission channel is established based on the Bluetooth protocol, this monitoring can be achieved through link layer signals in the Bluetooth communication protocol, such as detecting heartbeat packets or connection status codes, with the aim of promptly detecting whether the Bluetooth connection is normal.

[0020] S102: In response to the first connection state being disconnected, acquire the dialogue interruption data packet; the dialogue interruption data packet includes historical dialogue information within a first preset time period before the disconnection occurred, the user's voice information within a second preset time period after the disconnection occurred, and their respective timestamps.

[0021] In this embodiment, the conversation interruption data packet is a key data set collected by the wearable device when the Bluetooth connection is lost, used to resume the conversation. It includes historical conversation information before the Bluetooth connection was lost, user voice information after the Bluetooth connection was lost, and corresponding timestamps, ensuring that the terminal device can understand the complete conversation context. The first preset time period is a specific time window preset before the Bluetooth connection is lost, used to define the time range of historical conversation information that needs to be retained before the connection is lost, ensuring that the terminal device obtains sufficient conversation background. The second preset time period is a specific time window preset after the Bluetooth connection is lost, used to define the time range of user voice information that needs to be collected after the connection is lost, avoiding missing any content that the user continues to input after the connection is lost.

[0022] In this embodiment, when the first connection state is detected as disconnected, key dialogue data before and after the disconnection is quickly collected to ensure the integrity of the dialogue context. Specifically, this embodiment can collect two parts of information. The first part is historical dialogue information within a first preset time period before the disconnection, such as the previous reply from the terminal device or the previous command from the user, which is used by the terminal device to understand the dialogue context before the disconnection. The second part is the user's voice information within a second preset time period after the disconnection, such as the content that the user continues to input after the disconnection, to avoid the user repeating their expressions. At the same time, to ensure the temporal correlation of the information, corresponding timestamps are added to both parts of the information, such as the sending time of a certain voice message before the Bluetooth connection is disconnected and the collection time of the user's voice message after the Bluetooth connection is disconnected, which are finally integrated into a dialogue interruption data packet.

[0023] S103: In response to the first connection state failing to recover from the disconnected state to the connected state within a third preset time period, the second connection state of the second information transmission channel between the wearable device and the terminal device is detected; the second information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Wi-Fi connection, and the communication priority of the first information transmission channel is higher than the communication priority of the second information transmission channel.

[0024] In this embodiment, the third preset time period is a time window for Bluetooth to automatically resume after disconnection, used to determine whether the Bluetooth connection is momentary and avoid unnecessary channel switching. The second information transmission channel is a wireless communication channel established between the wearable device and the terminal device based on the Wi-Fi protocol. As a backup communication method after Bluetooth disconnection, it has a longer transmission distance and higher bandwidth, but relatively higher power consumption. The second connection status refers to the connection status of the second information transmission channel, including a connected status (i.e., the Wi-Fi link is normal and data can be transmitted) and a disconnected status (i.e., the Wi-Fi link is not established or is interrupted).

[0025] S104: In response to the second connection state being in the connection state, the information transmission channel between the wearable device and the terminal device is switched from the first information transmission channel to the second information transmission channel, and the dialogue interruption data packet is sent to the terminal device through the second information transmission channel so that the terminal device generates corresponding reply information based on the dialogue interruption data packet; In this embodiment, after Bluetooth is disconnected, the wearable device waits for a third preset time period to determine whether the Bluetooth connection can be automatically restored, as Bluetooth disconnection may be caused by momentary interference. If Bluetooth does not restore within the third preset time period, the device switches to checking the status of the second information transmission channel (based on Wi-Fi connection). Since the first information transmission channel (Bluetooth) has a higher priority, channel switching is only performed when Bluetooth cannot be restored and Wi-Fi is available, switching information transmission from Bluetooth to Wi-Fi. After the switch is complete, the interrupted data packet is sent to the terminal device via Wi-Fi to ensure that the terminal device can generate an accurate response based on the complete context.

[0026] S105: Receive the reply information from the terminal device through the second information transmission channel to restore the dialogue state between the wearable device and the terminal device.

[0027] In this embodiment, after the terminal device generates a response message based on the received conversation interruption data packet, it can send the response back to the wearable device through the switched Wi-Fi channel. Upon receiving the response, the wearable device can present it to the user through audio playback or display, allowing the user to continue with the next interaction based on the response. This achieves seamless resumption of the conversation state and avoids conversation termination or information loss due to disconnection.

[0028] For example, a user wears smart glasses while outdoors and connects to their phone via Bluetooth to speak with a voice assistant.

[0029] When the user walked into an area with strong signal interference, the Bluetooth connection was unexpectedly lost. At this time, the user was asking, "Where is the meeting location this afternoon? I remember it was at…". The smart glasses immediately detected that the Bluetooth connection had become disconnected.

[0030] The smart glasses retrieve historical conversation information from the local cache within a first preset time period (e.g., 3 seconds) before the conversation ended, specifically the voice data of "3 PM today" last sent by the phone; simultaneously, they continue to collect the user's subsequent voice messages within a second preset time period (e.g., 2 seconds) after the conversation ended, such as "...which conference room in the headquarters building?". All of this information is accompanied by a precise timestamp, collectively forming the conversation interruption data packet.

[0031] The smart glasses continuously attempted to reconnect via Bluetooth within a third preset time period (e.g., 2 seconds). Due to being in a signal dead zone, Bluetooth failed to automatically recover. The smart glasses then detected that the previously established Wi-Fi connection between the smart glasses and the phone was still available, so they performed a channel switch, sending the interrupted data packet to the phone via the Wi-Fi network.

[0032] After receiving the data packet, the voice assistant on the mobile phone, combined with the historical context of "3 PM today" and the new question "Which conference room in the headquarters building?", generated a coherent response: "Your meeting at 3 PM today is in conference room 1501 on the 15th floor of the headquarters building." This response was transmitted back to the smart glasses via the Wi-Fi channel and played to the user, successfully restoring the conversation that was about to be terminated due to Bluetooth interruption.

[0033] As can be seen from the above, this embodiment effectively solves the problem of dialogue content fragmentation that may occur when the connection is interrupted by creating and sending a dialogue interruption data packet that integrates key contextual information before and after the disconnection, thus ensuring semantic continuity. Simultaneously, by introducing a "wait-detect-switch" intelligent channel switching mechanism and prioritizing low-power Bluetooth, it not only avoids unnecessary switching caused by momentary interference, ensuring device energy efficiency, but also seamlessly enables Wi-Fi as a backup channel when Bluetooth cannot be restored, ensuring continuous communication capabilities. Ultimately, it overcomes the drawbacks of silent interruption handling in existing technologies, enabling intelligent, coherent, and reliable recovery of dialogue services between wearable devices and terminal devices after unexpected connection interruptions, greatly improving the user's interactive experience and service reliability in mobile scenarios.

[0034] In one embodiment of this application, the first connection state of the first information transmission channel includes a connected state and a disconnected state; During the dialogue between the wearable device and the terminal device, determining the first connection state of the first information transmission channel between the wearable device and the terminal device includes: During the dialogue between the wearable device and the terminal device, the communication quality parameters of the first information transmission channel are collected in real time; the communication quality parameters include at least two of the following: signal strength, bit error rate, and communication delay. If the communication quality parameters do not exceed their respective preset parameter threshold ranges within a consecutive preset number of detection periods, then the first connection state is determined to be a connection state. If, within a consecutive preset number of detection cycles, at least two of the communication quality parameters exceed their respective preset parameter threshold ranges, then the first connection state is determined to be a disconnected state.

[0035] In existing technologies, the connection or disconnection status is usually determined based on a single communication quality parameter, such as signal strength alone, or by directly determining the connection or disconnection status based on a single detection of a parameter exceeding a threshold. This lacks multi-parameter joint monitoring and continuous periodic verification, and is prone to misjudgment due to instantaneous fluctuations.

[0036] In this embodiment, communication quality parameters are quantitative indicators used to measure the communication performance and stability of the first information transmission channel. Specifically, these include signal strength, bit error rate, and communication latency. At least two of these parameters must be selected for joint monitoring to avoid the bias of a single indicator and ensure the accuracy of status judgment. A preset number of detection cycles is a fixed time interval for the wearable device to collect communication quality parameters, i.e., a detection cycle, such as 1 second / cycle. Data must be collected continuously for a preset number of cycles for judgment to filter out signal weakness caused by instantaneous fluctuations in parameters, avoiding erroneous operations that judge disconnection based on a single anomaly, and improving the robustness of status judgment. The preset parameter threshold range is a pre-configured numerical range in the wearable device used to determine whether the communication quality parameters are normal. Each parameter corresponds to an independent threshold. The preset parameter threshold range can be set based on the Bluetooth protocol standard and the actual usage scenario of the wearable device. Parameters within the range indicate that the link meets the communication requirements; parameters exceeding the range indicate a link anomaly. Among them, the real-time acquisition of communication quality parameters of the first information transmission channel emphasizes that the status of the first information transmission channel needs to be continuously tracked during the dialogue between the wearable device and the terminal device; while the detection cycle is a set fixed time interval, such as 1 second / cycle, so that the wearable device can periodically collect communication quality parameters according to this interval.

[0037] In this embodiment, during the dialogue between the wearable device and the terminal device, the wearable device can continuously collect communication quality parameters of the first information transmission channel, selecting at least two of the following: signal strength, bit error rate, and communication latency. Signal strength directly reflects the physical connection stability of the Bluetooth link; bit error rate reflects the accuracy of data transmission; and communication latency reflects the real-time performance of the link. Using at least two parameters is to avoid the limitations of a single parameter and improve the comprehensiveness of the status assessment.

[0038] If, within a predetermined number of consecutive detection periods, all collected communication quality parameters remain within their corresponding predetermined threshold ranges, the first connection state is determined to be a connected state. The fact that none exceed the thresholds indicates that the physical stability, data accuracy, and real-time interaction of the Bluetooth link meet the dialogue requirements, ensuring normal interaction. If, within a predetermined number of consecutive detection periods, any two of the collected communication quality parameters exceed their corresponding predetermined threshold ranges, the first connection state is determined to be a disconnected state. If two core parameters are abnormal, it indicates a substantial link failure, not just minor interference. Even if the third parameter is temporarily normal, the continuity of the dialogue cannot be guaranteed. In this case, a disconnected state is determined, which can promptly trigger subsequent processes such as dialogue interruption data packet collection, reducing user experience loss.

[0039] As can be seen from the above, this embodiment avoids misjudgment due to the one-sidedness of a single parameter and instantaneous fluctuations by collecting at least two communication quality parameters in real time and combining them with continuous detection cycle. The logic of determining a connection state when all parameters are within the preset parameter threshold range and determining a disconnect state when any two parameters exceed the preset parameter threshold range can accurately identify the true state of the Bluetooth channel, providing a reliable basis for subsequent disconnection processing, effectively reducing dialogue interference or delayed response caused by misjudgment, ensuring dialogue continuity, and improving the user interaction experience.

[0040] In one embodiment of this application, the first connection state of the first information transmission channel further includes an unstable connection state; During the dialogue between the wearable device and the terminal device, determining the first connection state of the first information transmission channel between the wearable device and the terminal device also includes: If, within a consecutive preset number of detection cycles, only one communication quality parameter exceeds the preset parameter threshold range corresponding to it, then the first connection state is determined to be an unstable connection state. The methods also include: If the first connection state is unstable, then a connection relationship is established between the wearable device and the terminal device through the second information transmission channel.

[0041] In this embodiment, the unstable connection state is an intermediate state between the connected state and the disconnected state, representing a partial but not complete failure of the first information transmission channel. Its trigger condition is that within a preset number of consecutive detection cycles, only one communication quality parameter exceeds a preset threshold. At this point, some data can still be transmitted, but the quality is insufficient for dialogue requirements. Its function is to provide early warning of Bluetooth channel failure risks and to trigger a pre-connection to a backup channel. The pre-connection process involves the wearable device performing basic link establishment operations for the second information transmission channel when the first information transmission channel is in an unstable connection state. This includes completing low-level preparations such as Wi-Fi protocol negotiation, authentication, and link layer synchronization, but without transmitting upper-layer dialogue data. This shortens the channel switching time during subsequent disconnections, avoids delays caused by establishing a Wi-Fi connection from scratch, and does not interfere with the normal communication of the current Bluetooth channel.

[0042] In this embodiment, if only one of the collected communication quality parameters exceeds the corresponding preset parameter threshold range within a consecutive preset number of detection cycles, the first connection state is determined to be an unstable connection state. Setting any parameter exceeding the threshold as the trigger condition is because in this scenario, although the Bluetooth channel is not completely disconnected, a partial fault has occurred, such as high communication latency causing dialogue stuttering or high bit error rate causing voice segment loss. Without early intervention, it is highly likely to develop into a disconnected state.

[0043] Once an unstable connection is detected, the wearable device can immediately initiate a pre-connection process for the second information transmission channel. This establishes the underlying link required for communication, but does not transmit conversation data. The key reason for this design is to avoid interfering with the current Bluetooth channel communication. Because the first information transmission channel has higher priority, the pre-connection only serves as basic preparation, without preempting bandwidth or switching data transmission paths, thus avoiding impacting the real-time performance of the current conversation. It also shortens subsequent switching latency. If the unstable connection later develops into a disconnection, there's no need to rebuild the Wi-Fi connection from scratch; conversation interruption data packets can be directly transmitted based on the established underlying link, significantly reducing the waiting time after disconnection and minimizing perceived lag for the user.

[0044] As can be seen from the above, in traditional logic, Bluetooth only starts detecting and establishing a Wi-Fi connection after it suddenly disconnects from a normal connection, resulting in a significant window period. However, this embodiment transforms the passive response of disconnection-switching into an active response of warning-preparation-rapid switching by providing early warnings for unstable states and preparing for pre-connection in advance, thereby further ensuring the seamlessness of the conversation.

[0045] In one embodiment of this application, it further includes: Based on historical dialogue data stored in wearable devices, the dialogue frequency characteristics between wearable devices and terminal devices and the user's speech rate characteristics are obtained. The dialogue frequency characteristics are the average time interval between the start times of two adjacent voice commands initiated by the user, calculated based on historical dialogue data, and the user's speech rate characteristics are the average speech rate of the user in historical dialogue data. The methods for determining the first preset time period include: In response to a continuous interaction interval in historical dialogue data being less than a preset interval threshold, the basic preset time is shortened based on a first adjustment step size to determine a first preset time period; or in response to a continuous interaction interval in historical dialogue data being greater than or equal to a preset interval threshold, the basic preset time is extended based on a first adjustment step size to determine a first preset time period. The methods for determining the second preset time period include: In response to the user's average speaking speed being greater than a preset speaking speed threshold, the basic preset time is extended based on a second adjustment step size to determine a second preset time period; or in response to the user's average speaking speed being less than or equal to a preset speaking speed threshold, the basic preset time is shortened based on a second adjustment step size to determine a second preset time period.

[0046] In this embodiment, historical dialogue data refers to the dialogue logs of all successful interactions between the user and the terminal device over a past period, recorded in the wearable device's local cache. This log includes dialogue content, interaction timestamps, and the temporal relationship between commands and responses. Dialogue frequency characteristics are an indicator reflecting the intensity of user-device interaction, specifically manifested as the continuous interaction interval. This is the average time interval between the end of one user's voice command and the start of the next in historical dialogues. For example, if a user has three consecutive conversations with the smart glasses, the first voice command ends at 9:00:00, the second begins at 9:00:03 (3 seconds apart), the second ends at 9:00:07, and the third begins at 9:00:09 (2 seconds apart), resulting in an average interval of (3+2) / 2 = 2.5 seconds. A shorter interval indicates a higher dialogue frequency.

[0047] User speech rate characteristics are indicators reflecting the speed of a user's expression, specifically manifested as average speech rate. This is calculated by analyzing historical speech data and determining the number of syllables or characters uttered by the user per unit of time; a higher value indicates a faster speaking speed. The preset interval threshold is a pre-configured baseline time value in the wearable device used to determine whether conversations are frequent or sparse. It can be set based on typical usage scenarios of the wearable device (such as casual conversation or command queries) and the average interaction habits of most users. The preset speech rate threshold is a pre-configured baseline speed value in the wearable device used to determine whether a user's speech rate is fast or slow; it can typically be set with reference to the average speech rate of adults in daily conversation. The basic preset time is a pre-configured initial baseline time value in the wearable device used to dynamically adjust the first and second preset time periods; it can be set based on the memory size configured in the wearable device.

[0048] In this embodiment, the frequency features of the dialogue and the user's speech rate features are first extracted from the historical dialogue data stored in the wearable device. This provides a basis for dynamically determining the first and second preset time periods, avoiding the use of a uniform and fixed time window to adapt to all users, which could lead to data redundancy or omission of key information for some users.

[0049] In this embodiment, if the interval between consecutive interactions in the historical dialogue data is less than a preset interval threshold, the basic preset time is shortened based on the first adjustment step size to determine the first preset time period. For example, if the basic preset time is 30 seconds, after shortening it by 20%, the first preset time period is 24 seconds. This can avoid collecting too many repetitive or irrelevant historical dialogues and reduce the storage footprint of the wearable device.

[0050] If the interval between consecutive interactions in historical dialogue data is greater than or equal to a preset interval threshold, the basic preset time is extended based on the first adjustment step to determine the first preset time period. For example, if the basic preset time is 30 seconds, after extending it by 20%, the first preset time period is 36 seconds, which can ensure that key dialogues before the disconnection are covered and avoid context loss.

[0051] In this embodiment, if the user's average speaking speed exceeds a preset speaking speed threshold, the basic preset time is extended based on a second adjustment step size to determine a second preset time period. For example, if the basic preset time is 15 seconds, after extending it by 15%, the second preset time period is 17.25 seconds. This avoids the recording stopping before the voice is finished due to a disconnection caused by the user's fast speaking speed.

[0052] If the user's average speaking speed is less than or equal to the preset speaking speed threshold, the basic preset time is shortened based on the second adjustment step size to determine the second preset time period. For example, if the basic preset time is 15 seconds, after shortening it by 15%, the second preset time period is 12.75 seconds. This avoids collecting too much silent or repetitive content and optimizes the utilization of storage and transmission resources.

[0053] In this embodiment, determining the first preset time period based on historical dialogue data is to adapt to the rhythm of the dialogue, and determining the second preset time period based on the user's average speaking speed is to adapt to the expression speed. The purpose is to make the time window fit the user's habits, balance the integrity of the context and the efficiency of resources, and provide high-quality data for dialogue recovery.

[0054] As can be seen from the above, this embodiment dynamically determines a first preset time period and a second preset time period by extracting dialogue frequency and user speech rate characteristics from historical dialogue data: the first preset time period is shortened when dialogue is dense to prevent data redundancy, and extended when dialogue is sparse to maintain context integrity; the second preset time period is extended when the user speaks quickly to prevent missed data collection, and shortened when speaking slowly to reduce useless information. This approach adapts to user habits, optimizes device storage and transmission resources, ensures dialogue data quality, provides reliable support for subsequent dialogue recovery, and improves the interactive experience.

[0055] In one embodiment of this application, after sending the conversation interruption data packet to the terminal device through the second information transmission channel, the method further includes: If no acknowledgment signal is received within the third preset time, it is determined that the conversation interruption data packet transmission failed. The acknowledgment signal is the signal returned by the terminal device after receiving the conversation interruption data packet. In response to a failed data packet transmission due to a dialogue interruption, the network bandwidth of the current second information transmission channel is collected in real time. The interruption packets are processed based on network bandwidth, and the processed interruption packets are retransmitted.

[0056] In this embodiment, the acknowledgment signal is a response signal returned by the terminal device to the wearable device after successfully receiving the interrupted data packet sent by the wearable device, used to inform the wearable device that the data packet has been completely received. The third preset time period is the maximum duration for the wearable device to wait for the terminal device to return the acknowledgment signal after sending the interrupted data packet, which is pre-configured by the wearable device. Network bandwidth is the amount of data that the second information transmission channel can transmit per unit time, reflecting the actual data carrying capacity of the current network.

[0057] In this embodiment, after sending the conversation interruption data packet through the second information transmission channel, the wearable device waits for an acknowledgment signal from the terminal device, and the waiting time is limited to a third preset time, such as 5 seconds. The acknowledgment signal is a response information generated by the terminal device after successfully receiving the data packet. If received, it indicates successful transmission, and the subsequent process of receiving a reply can begin; if not received, a failure handling mechanism is triggered to avoid conversation recovery stalling due to silent failure. If no acknowledgment signal is received from the terminal device within the third preset time, the wearable device can determine that the conversation interruption data packet transmission failed, avoiding indefinite waiting and ensuring timely initiation of remedial measures.

[0058] Upon determining that the interruption packet transmission failed, the wearable device immediately collects the network bandwidth of the second information transmission channel in real time. If the bandwidth is too low, it may be due to a transmission timeout caused by an excessively large data packet; if the bandwidth is normal, it may be due to packet loss caused by momentary interference. Based on the collected network bandwidth, the interruption packet is processed and retransmitted to ensure that the interruption packet eventually reaches the terminal device, providing a data foundation for subsequent conversation recovery.

[0059] As can be seen from the above, this embodiment promptly detects data packet transmission failures through a confirmation signal mechanism, avoiding silence failures that could lead to a halt in dialogue recovery. Combined with real-time network bandwidth processing and retransmission of data packets, it adapts to different Wi-Fi conditions and improves the retransmission success rate. This effectively ensures the reliable transmission of data packets during dialogue interruptions, providing complete data for the terminal to generate responses, ensuring the continuity of dialogue state recovery, and optimizing the user experience.

[0060] In one embodiment of this application, processing of session interruption data packets based on network bandwidth and retransmission of the processed session interruption data packets includes: In response to network bandwidth being within a first preset bandwidth range, the session interruption data packet is divided into multiple sub-data packets, and the multiple sub-data packets are sorted. Based on the sorting result, each sub-data packet is transmitted sequentially. In response to network bandwidth being within the second preset bandwidth range, the interrupted data packet is compressed to obtain a compressed data packet, and the compressed data packet is retransmitted. The upper limit of the first preset bandwidth range is less than the lower limit of the second preset bandwidth range.

[0061] In this embodiment, the first preset bandwidth range is a fixed interval adapted to extremely low bandwidth scenarios. When the real-time collected network bandwidth is less than the first preset bandwidth range, fragmentation processing is triggered. First, the original conversation interruption data packet is divided into multiple independent sub-data packets according to a preset size (e.g., 50KB / packet) to ensure that the size of a single sub-data packet can adapt to the current extremely low bandwidth carrying capacity and avoid packet loss due to excessive pressure in a single transmission. Next, a unique sorting identifier (e.g., "1 / 3, 2 / 3, 3 / 3") is added to all sub-data packets to clarify the positional relationship of each sub-data packet in the original data packet. Finally, based on the sorting result, each sub-data packet is transmitted sequentially.

[0062] This breaks down the high-pressure operation of transmitting large data packets at once into low-pressure operations of transmitting small data packets multiple times, significantly reducing the risk of packet loss in a single transmission. If a sub-data packet fails to be transmitted, only that sub-data packet needs to be retransmitted, without having to retransmit the entire original data packet, reducing the amount of data to be retransmitted and further adapting to resource constraints with extremely low bandwidth.

[0063] In this embodiment, the second preset bandwidth range is a fixed interval adapted to medium bandwidth scenarios, and its lower limit is equal to the upper limit of the first preset bandwidth range, ensuring that the bandwidth intervals of the two strategies are seamlessly connected and do not overlap.

[0064] When the real-time network bandwidth is within this range, compression processing is triggered. A lightweight data compression algorithm is used. While ensuring the integrity of the dialogue information, the size of the dialogue interruption data packet is reduced; after compression, a compressed data packet is generated, and the compressed packet is directly retransmitted without additional fragmentation.

[0065] By reducing the data size through compression, data packets can be transmitted quickly under medium bandwidth, avoiding transmission timeouts caused by excessive data volume. Compared to fragmentation, compression does not require splitting data packets or adding sorting identifiers, making the processing flow simpler, retransmission efficiency higher, and balancing bandwidth adaptability and transmission timeliness.

[0066] The upper limit of the first preset bandwidth range is equal to the lower limit of the second preset bandwidth range. The two bandwidth ranges are seamlessly connected and have no overlap, ensuring that any real-time bandwidth can only be matched with one processing strategy, avoiding logical conflicts. For example: if the bandwidth is 0.8Mbps and only falls within the first preset range (0-1Mbps), only fragmentation will be triggered; if the bandwidth is 1.5Mbps and only falls within the second preset range (1-3Mbps), only compression will be triggered, and no additional priority judgment is required.

[0067] For example, when a user wears smart glasses and connects to a mobile phone via Wi-Fi, if the Bluetooth connection is lost, the smart glasses will send a 150KB data packet containing the questions "Is it raining today?" and "Do I need an umbrella?" to the mobile phone, and then wait for a 5-second (third preset time) confirmation signal. If no confirmation is received from the mobile phone after 5 seconds, the transmission is considered to have failed, and the current Wi-Fi bandwidth is immediately collected. If the detected bandwidth is 0.8Mbps (within the first preset bandwidth range of 0-1Mbps), the smart glasses trigger fragmentation processing: the 150KB data packet is split into three 50KB sub-data packets, sorted by a sequencer, and transmitted sequentially. Upon receiving the data, the mobile phone reassembles the data packets according to the sequencer, returns an acknowledgment signal, and the retransmission is successful.

[0068] If the detected bandwidth is 1.5Mbps (within the second preset bandwidth range of 1-3Mbps), compression processing is triggered: the 150KB data packet is compressed to 90KB using the OPUS algorithm, and the compressed packet is retransmitted directly. After the mobile phone decompresses the packet, it verifies its integrity, returns a confirmation signal, and the retransmission is successful.

[0069] Both processing methods ensure reliable data packet transmission. The mobile phone generates a reply such as "It will rain lightly today, we recommend bringing a folding umbrella" based on the complete data packet, and sends the reply to the smart glasses via Wi-Fi to resume the weather query conversation.

[0070] As can be seen from the above, this embodiment divides the bandwidth into a first and a second preset bandwidth range that do not overlap, and then employs fragmentation and sorting transmission and compression retransmission strategies within the corresponding ranges: fragmentation reduces transmission pressure and packet loss when bandwidth is low; compression reduces data volume and improves efficiency when bandwidth is high. The non-overlapping ranges avoid logical conflicts, accurately adapt to network conditions, improve retransmission success rate, ensure reliable transmission of interrupted data packets, and optimize the user experience.

[0071] In one embodiment of this application, after detecting the second connection status of the second information transmission channel between the wearable device and the terminal device, the method further includes: In response to the second information transmission channel being disconnected, the system parses the interrupted data packet based on the target model and generates a local response message; the target model is a trained lightweight interaction model set in the wearable device.

[0072] In this embodiment, the target model is a lightweight, trained interaction model deployed locally on the wearable device, possessing basic dialogue parsing and response generation capabilities. This target model has few parameters and low computational requirements, requiring no reliance on the cloud or terminal devices; it can run in real-time using only the wearable device's low-power processor, perfectly adapting to the limited storage and computing resources of such devices. The target model possesses basic dialogue parsing and intent recognition capabilities, able to extract the user's core needs (such as "Do you need an umbrella?") and incomplete context from dialogue interruption data packets, generating context-appropriate temporary responses (such as "Your request has been received; results will be synchronized immediately after network recovery"). However, it cannot perform complex semantic understanding or in-depth explanations. This differs significantly from terminal device responses: terminal device responses rely on a connection channel, generating formal and accurate in-depth explanations based on complete data and stronger computing power (such as "It's drizzling today; we recommend bringing a folding umbrella"), while the target model's response is a transitional response during a complete disconnection, intended to avoid user silence and merely confirm receipt and inform the user of the status. The local response information is the temporary response content generated by the target model based on dialogue interruption data packets, used to respond to user needs when the entire channel is disconnected.

[0073] In this embodiment, when the second information transmission channel is also detected to be disconnected (i.e., Bluetooth and Wi-Fi cannot connect to the terminal device), the wearable device no longer waits for the external channel to be restored, but instead inputs the previously collected conversation interruption data packets into the target model. The target model, as a lightweight interaction model deployed locally on the wearable device, can quickly parse data packets based on limited computing power, identify incomplete context from historical conversations or new intentions after the user disconnects, and generate local response information, thus avoiding the user facing an unresponsive silence.

[0074] The local response is not a formal reply from the terminal device, but a transitional response from the wearable device in a completely disconnected scenario. It can both confirm to the user that the request has been received and maintain the continuity of the conversation, solving the problem that the conversation ends when the entire channel is disconnected in traditional technologies.

[0075] In this embodiment, during the generation of local response information, the wearable device continuously monitors the connection status of the first and second information transmission channels in parallel. This ensures that once either channel regains connection, the data synchronization process is triggered immediately, preventing local information from becoming disconnected from the terminal device. When either information transmission channel is detected to have regained connection, the wearable device can send the original conversation interruption data packet and local response information to the terminal device through that channel. The terminal device generates a formal response based on these two pieces of data and then sends it back to the wearable device through the restored channel. This achieves a seamless transition from the local temporary response to the terminal's formal response, resulting in a complete restoration of the conversation state.

[0076] For example, when a user wears smart glasses and asks "the shortest route from home to the nearest fresh food supermarket" via voice, the Bluetooth connection between the glasses and the phone disconnects because the distance exceeds 10 meters. The glasses then check the Wi-Fi according to the procedure, find that the user is not connected to public Wi-Fi and the home Wi-Fi signal is too weak to be connected, triggering the local response mechanism.

[0077] The glasses input a conversation interruption data packet containing "route query voice and interaction records of the 30 seconds before disconnection" into the local target model. The model quickly parses the intent that "the user has not completed the route query" and plays a local response through the bone conduction speaker: "Your route query request has been received. The network is currently disconnected. We will continue to monitor the signal and update the route for you as soon as it is restored," thus avoiding the user from asking the same question repeatedly due to lack of response.

[0078] The glasses then monitor Bluetooth and Wi-Fi status every 2 seconds. When the user walks to the entrance of the community 50 meters away, the Wi-Fi reconnects to the public network of nearby shops. The glasses immediately send the interrupted conversation data packet and local reply record to the mobile phone via Wi-Fi. Based on the complete data packet, the mobile phone generates a formal route reply: "Go straight for 300 meters along the east gate of the community, then turn left. The whole journey takes 5 minutes." After being fed back to the glasses, the user can continue to receive navigation guidance, achieving a seamless resumption of the conversation.

[0079] As can be seen from the above, this embodiment generates a local response through a local lightweight target model when both Bluetooth and Wi-Fi are disconnected, responding to incomplete conversations or new user intentions, thus avoiding a silent state for the user. Simultaneously, it continuously monitors the channel, synchronizing data packets and local responses upon recovery to ensure the terminal generates accurate replies. This effectively solves the problem of conversation termination when all channels are disconnected, maintains interaction continuity, and improves the user experience.

[0080] Based on the same inventive concept, this application also provides a dialogue state recovery device for implementing the dialogue state recovery method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more dialogue state recovery device embodiments provided below can be found in the limitations of the dialogue state recovery method described above, and will not be repeated here.

[0081] This application provides a dialog state restoration device, such as... Figure 2 As shown, device 20 is installed in a wearable device. Device 20 includes: a first connection status determination module 21, a dialogue interruption data acquisition module 22, a second connection status determination module 23, a transmission channel switching module 24, and a dialogue status recovery module 25.

[0082] The first connection status determination module 21 is used to determine the first connection status of the first information transmission channel between the wearable device and the terminal device during the dialogue between the wearable device and the terminal device; the first information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Bluetooth connection. The dialogue interruption data acquisition module 22 is used to acquire a dialogue interruption data packet in response to the first connection state being disconnected. The dialogue interruption data packet includes historical dialogue information within a first preset time period before the disconnection occurred, the user's voice information within a second preset time period after the disconnection occurred, and their respective timestamps. The second connection status determination module 23 is used to detect the second connection status of the second information transmission channel between the wearable device and the terminal device in response to the first connection status failing to recover from the disconnected state to the connected state within a third preset time period; the second information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Wi-Fi connection, and the communication priority of the first information transmission channel is higher than the communication priority of the second information transmission channel. The transmission channel switching module 24 is used to switch the information transmission channel between the wearable device and the terminal device from the first information transmission channel to the second information transmission channel in response to the second connection state being in the connection state, and send the dialogue interruption data packet to the terminal device through the second information transmission channel so that the terminal device can generate corresponding reply information based on the dialogue interruption data packet; The dialogue state recovery module 25 is used to receive the reply information fed back by the terminal device through the second information transmission channel, so as to restore the dialogue state between the wearable device and the terminal device.

[0083] In one embodiment of this application, the first connection state of the first information transmission channel includes a connected state and a disconnected state; When determining the first connection state of the first information transmission channel between the wearable device and the terminal device, the first connection state determination module 21 is specifically used for: During the dialogue between the wearable device and the terminal device, the communication quality parameters of the first information transmission channel are collected in real time; the communication quality parameters include at least two of the following: signal strength, bit error rate, and communication delay. If the communication quality parameters do not exceed their respective preset parameter threshold ranges within a consecutive preset number of detection periods, then the first connection state is determined to be a connection state. If, within a consecutive preset number of detection cycles, at least two of the communication quality parameters exceed their respective preset parameter threshold ranges, then the first connection state is determined to be a disconnected state.

[0084] In one embodiment of this application, the first connection state of the first information transmission channel further includes an unstable connection state; When determining the first connection state of the first information transmission channel between the wearable device and the terminal device, the first connection state determination module 21 is further configured to: If, within a consecutive preset number of detection cycles, only one communication quality parameter exceeds the preset parameter threshold range corresponding to it, then the first connection state is determined to be an unstable connection state. If the first connection state is unstable, then a connection relationship is established between the wearable device and the terminal device through the second information transmission channel.

[0085] In one embodiment of this application, the device 20 further includes: The dialogue feature acquisition module is specifically used for: Based on historical dialogue data stored in wearable devices, the dialogue frequency characteristics between wearable devices and terminal devices and the user's speech rate characteristics are obtained. The dialogue frequency characteristics are the average time interval between the start times of two adjacent voice commands initiated by the user, calculated based on historical dialogue data, and the user's speech rate characteristics are the average speech rate of the user in historical dialogue data. The first preset time period determination module is specifically used for: In response to a continuous interaction interval in historical dialogue data being less than a preset interval threshold, the basic preset time is shortened based on a first adjustment step size to determine a first preset time period; or in response to a continuous interaction interval in historical dialogue data being greater than or equal to a preset interval threshold, the basic preset time is extended based on a first adjustment step size to determine a first preset time period. The second preset time period determination module is specifically used for: In response to the user's average speaking speed being greater than a preset speaking speed threshold, the basic preset time is extended based on a second adjustment step size to determine a second preset time period; or in response to the user's average speaking speed being less than or equal to a preset speaking speed threshold, the basic preset time is shortened based on a second adjustment step size to determine a second preset time period.

[0086] In one embodiment of this application, the device 20 further includes a signal confirmation module, which, after sending the conversation interruption data packet to the terminal device through the second information transmission channel, is specifically used for: If no acknowledgment signal is received within the third preset time, it is determined that the conversation interruption data packet transmission failed. The acknowledgment signal is the signal returned by the terminal device after receiving the conversation interruption data packet. In response to a failed data packet transmission due to a dialogue interruption, the network bandwidth of the current second information transmission channel is collected in real time. The interruption packets are processed based on network bandwidth, and the processed interruption packets are retransmitted.

[0087] In one embodiment of this application, when processing the interrupted session data packet according to the network bandwidth and retransmitting the processed interrupted session data packet, the signal confirmation module is further configured to: In response to network bandwidth being within a first preset bandwidth range, the session interruption data packet is divided into multiple sub-data packets, and the multiple sub-data packets are sorted. Based on the sorting result, each sub-data packet is transmitted sequentially. In response to network bandwidth being within the second preset bandwidth range, the interrupted data packet is compressed to obtain a compressed data packet, and the compressed data packet is retransmitted. The upper limit of the first preset bandwidth range is less than the lower limit of the second preset bandwidth range.

[0088] In one embodiment of this application, the device 20 further includes: a local response module, which, after detecting the second connection status of the second information transmission channel between the wearable device and the terminal device, is specifically used for: In response to the second information transmission channel being disconnected, the system parses the interrupted data packet based on the target model and generates a local response message; the target model is a trained lightweight interaction model set in the wearable device.

[0089] See Figure 3 , Figure 3 This is a schematic block diagram of a wearable device provided in one embodiment of this application. Figure 3 The wearable device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of each module / unit in the above-described device embodiments, for example... Figure 2 The functions of the first connection status judgment module 21, the dialogue interruption data acquisition module 22, the second connection status judgment module 23, the transmission channel switching module 24, and the dialogue status recovery module 25 are shown.

[0090] It should be understood that, in the embodiments of this application, the processor 301 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.

[0091] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0092] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store information such as conversation interruption packets and historical conversation data.

[0093] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation method described in the dialogue state recovery method provided in the embodiments of this application, or they can execute the implementation method of the wearable device described in the embodiments of this application, which will not be repeated here.

[0094] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0095] The computer-readable storage medium can be an internal storage unit of the wearable device in any of the foregoing embodiments, such as a hard drive or memory of the wearable device. The computer-readable storage medium can also be an external storage device of the wearable device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., provided on the wearable device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the wearable device. The computer-readable storage medium is used to store computer programs and other programs and data required by the wearable device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0096] Those skilled in the art will recognize that the modules / 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.

[0097] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the wearable device and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0098] In the several embodiments provided in this application, it should be understood that the disclosed wearable devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules, units, or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces or modules / units, or it may be an electrical, mechanical, or other form of connection.

[0099] The modules / units described as separate components may or may not be physically separate. Similarly, the components shown as modules / units may or may not be physical modules / units; they may be located in one place or distributed across multiple network modules / units. Some or all of the modules / units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0100] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The integrated modules described above can be implemented in hardware or in the form of software functional modules / units.

[0101] The above are merely specific embodiments 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 restoring a dialogue state, characterized in that, Applications in wearable devices, including: During the dialogue between the wearable device and the terminal device, the first connection state of the first information transmission channel between the wearable device and the terminal device is determined; the first information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Bluetooth connection. In response to the first connection state being disconnected, a conversation interruption data packet is obtained; the conversation interruption data packet includes historical conversation information within a first preset time period before the disconnection occurred, the user's voice information within a second preset time period after the disconnection occurred, and their respective timestamps. In response to the first connection state failing to recover from the disconnected state to the connected state within a third preset time period, the second connection state of the second information transmission channel between the wearable device and the terminal device is detected; the second information transmission channel is a signal transmission channel between the wearable device and the terminal device based on a Wi-Fi connection, and the communication priority of the first information transmission channel is higher than the communication priority of the second information transmission channel; In response to the second connection state being in a connected state, the information transmission channel between the wearable device and the terminal device is switched from the first information transmission channel to the second information transmission channel, and the conversation interruption data packet is sent to the terminal device through the second information transmission channel, so that the terminal device generates corresponding reply information based on the conversation interruption data packet; The device receives the response information from the terminal device through the second information transmission channel to restore the dialogue state between the wearable device and the terminal device.

2. The dialogue state restoration method as described in claim 1, characterized in that, The first connection state of the first information transmission channel includes a connected state and a disconnected state; During the dialogue between the wearable device and the terminal device, determining the first connection state of the first information transmission channel between the wearable device and the terminal device includes: During the dialogue between the wearable device and the terminal device, the communication quality parameters of the first information transmission channel are collected in real time; the communication quality parameters include at least two of signal strength, bit error rate and communication delay. If the communication quality parameters do not exceed their respective preset parameter threshold ranges within a consecutive preset number of detection periods, then the first connection state is determined to be a connection state. If, within a consecutive preset number of detection cycles, at least two of the communication quality parameters exceed their respective preset parameter threshold ranges, then the first connection state is determined to be a disconnected state.

3. The dialogue state restoration method as described in claim 2, characterized in that, The first connection state of the first information transmission channel also includes an unstable connection state; The step of determining the first connection state of the first information transmission channel between the wearable device and the terminal device during the dialogue between the wearable device and the terminal device further includes: If, within a consecutive preset number of detection cycles, only one of the communication quality parameters exceeds the preset parameter threshold range corresponding to it, then the first connection state is determined to be an unstable connection state. The method further includes: If the first connection state is an unstable connection state, then a connection relationship is established between the wearable device and the terminal device through the second information transmission channel.

4. The dialogue state restoration method as described in claim 1, characterized in that, Also includes: Based on the historical dialogue data stored in the wearable device, the dialogue frequency characteristics and user speech rate characteristics between the wearable device and the terminal device are obtained. The dialogue frequency feature is the average time interval between the start times of two consecutive voice commands initiated by the user in the dialogue, calculated based on the historical dialogue data, and the user's speech rate feature is the average speech rate of the user in the historical dialogue data. The method for determining the first preset time period includes: In response to the continuous interaction interval in the historical dialogue data being less than a preset interval threshold, the basic preset time is shortened based on a first adjustment step size to determine the first preset time period; or in response to the continuous interaction interval in the historical dialogue data being greater than or equal to a preset interval threshold, the basic preset time is extended based on a first adjustment step size to determine the first preset time period. The method for determining the second preset time period includes: In response to the user's average speaking speed being greater than a preset speaking speed threshold, the basic preset time is extended based on a second adjustment step size to determine the second preset time period; or in response to the user's average speaking speed being less than or equal to a preset speaking speed threshold, the basic preset time is shortened based on a second adjustment step size to determine the second preset time period.

5. The dialogue state restoration method as described in claim 1, characterized in that, After sending the conversation interruption data packet to the terminal device through the second information transmission channel, the method further includes: If no acknowledgment signal is received within the third preset time, it is determined that the transmission of the conversation interruption data packet has failed. The acknowledgment signal is the signal returned by the terminal device after receiving the conversation interruption data packet. In response to the failure to send the interrupted data packet, the network bandwidth of the second information transmission channel is collected in real time. The session interruption data packet is processed according to the network bandwidth, and the processed session interruption data packet is retransmitted.

6. The dialogue state restoration method as described in claim 5, characterized in that, The step of processing the session interruption data packet according to the network bandwidth and retransmitting the processed session interruption data packet includes: In response to the network bandwidth being within a first preset bandwidth range, the conversation interruption data packet is divided into multiple sub-data packets, and the multiple sub-data packets are sorted. Based on the sorting result, each sub-data packet is transmitted sequentially. In response to the network bandwidth being within a second preset bandwidth range, the interrupted data packet is compressed to obtain a compressed data packet, and the compressed data packet is retransmitted. The upper limit of the first preset bandwidth range is less than the lower limit of the second preset bandwidth range.

7. The dialogue state restoration method as described in claim 1, characterized in that, After detecting the second connection status of the second information transmission channel between the wearable device and the terminal device, the method further includes: In response to the second information transmission channel being disconnected, the local response information is generated by parsing the dialogue interruption data packet based on the target model; the target model is a trained lightweight interaction model set in the wearable device.

8. A dialogue state restoration device, characterized in that, The device is disposed in a wearable device, the device comprising: The first connection status determination module is used to determine the first connection status of the first information transmission channel between the wearable device and the terminal device during the dialogue between the wearable device and the terminal device; the first information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Bluetooth connection. The dialogue interruption data acquisition module is used to acquire a dialogue interruption data packet in response to the first connection state being disconnected; the dialogue interruption data packet includes historical dialogue information within a first preset time period before the disconnection state occurred, the user's voice information within a second preset time period after the disconnection state occurred, and their respective timestamps. The second connection status determination module is used to detect the second connection status of the second information transmission channel between the wearable device and the terminal device in response to the first connection status failing to recover from the disconnected state to the connected state within a third preset time period; the second information transmission channel is a signal transmission channel between the wearable device and the terminal device based on Wi-Fi connection, and the communication priority of the first information transmission channel is higher than the communication priority of the second information transmission channel; The transmission channel switching module is used to switch the information transmission channel between the wearable device and the terminal device from the first information transmission channel to the second information transmission channel in response to the second connection state being in a connected state, and send the dialogue interruption data packet to the terminal device through the second information transmission channel, so that the terminal device generates corresponding reply information based on the dialogue interruption data packet; The dialogue state recovery module is used to receive the reply information fed back by the terminal device through the second information transmission channel, so as to restore the dialogue state between the wearable device and the terminal device.

9. A wearable device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images