Road condition information prompting method and device based on machine learning and earphone

Abstract

The application discloses a road condition information prompting method and device based on machine learning and earphones. The method comprises the following steps: S100, acquiring a static posture parameter when the earphones are worn; S200, collecting a dynamic parameter of a user's head by a gravity sensor when the user moves along a preset route; S300, iteratively learning a behavior posture of the user during movement along the preset route according to the dynamic parameter, so as to record the user's movement habit along the preset route; S400, marking a node to be reminded in the preset route; and S500, sending first prompt information to the user when the user moves to the node to be reminded along the preset route, so as to remind the user to pay attention to road condition information. Thus, the road condition information can be effectively prompted to the user while minimizing the experience of using the earphones.

Description

Technical Field

[0001] The present invention relates to the field of audio processing technology, and in particular to a method, device and headset for providing road condition information based on machine learning. Background Art

[0002] Currently, headphones are one of the hottest-selling consumer electronics products, often carried around for listening to music, making calls, and more. Headphones are widely used in daily life, and students and office workers often wear them when traveling. Their routes to school or work are often fixed, and as they travel more frequently, they lose focus on road conditions. This can lead to inattention to changes in road conditions, such as when passing crosswalks or intersections. Furthermore, existing headphones utilize active noise cancellation, which is effective in rejecting most external noise signals. However, when users wear these headphones to listen to music or make calls, they often become immersed in their own world, unable to quickly access information about road conditions, leading to bumps and accidents.

[0003] To avoid bumps, collisions, traffic accidents, etc., users can turn off the noise reduction function of the headphones (i.e. switch to transparency mode). On the one hand, this requires users to actively turn on or off the noise reduction function, which is cumbersome for users; on the other hand, when the noise reduction function is turned off, users will hear noisy ambient noise, which affects the user experience and is not suitable for long-term use.

[0004] Therefore, when a user uses headphones, how to effectively prompt the user with road condition information while minimizing the impact on the user's experience of using headphones has become a technical problem that needs to be solved urgently. Summary of the Invention

[0005] Based on the above situation, the main purpose of the present invention is to provide a road condition information prompting method, device and headphones based on machine learning, so as to effectively prompt the user with road condition information while minimizing the user's experience of using headphones when the user uses the headphones.

[0006] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, an embodiment of the present invention discloses a method for providing road condition information based on machine learning, comprising:

[0008] Step S100, obtaining static posture parameters of the headset when being worn, the static posture parameters including a first static parameter and a second static parameter, wherein the first static parameter is a gravity sensor posture parameter corresponding to a state in which the user is looking forward while wearing the headset, and the second static parameter is a gravity sensor posture parameter corresponding to a state in which the user is looking down while wearing the headset;

[0009] Step S200: When the user moves along a preset route, the dynamic posture of the user's head is collected by the gravity sensor to obtain dynamic parameters. The dynamic parameters are the parameters of the gravity sensor relative to the static posture parameters when the user wears the headset. The preset route is the route the user moves from the starting point to the destination.

[0010] Step S300, iteratively learning the user's behavior and posture during movement along the preset route based on the dynamic parameters to record the user's movement habits along the preset route;

[0011] Step S400: Marking nodes to be reminded in the preset route, where the nodes to be reminded include nodes whose corresponding dynamic parameters vary by more than a threshold relative to the static posture parameters;

[0012] Step S500: When the user moves to a node to be reminded along a preset route, a first reminder message is sent to the user to remind the user to pay attention to road condition information.

[0013] Optionally, in step S500, the first prompt information is a specific audio prompt tone.

[0014] Optionally, before step S400, the method further includes:

[0015] Step S210, when the user moves along the preset route, the microphone collects ambient sound;

[0016] Step S220, identifying ambient sounds. When the target sound category is detected in the ambient sounds, step S410 is executed.

[0017] Step S410: Marking a target environment node in the preset route, where the target environment node is a node in the preset route where sounds of the target category exist;

[0018] Step S500 further includes:

[0019] When the user moves to the target environment node along the preset route, a second prompt message is sent to the user to remind the user to pay attention to the current environment.

[0020] Optionally, in step S500, sending the second prompt information to the user includes:

[0021] Step S510: extracting target-category audio from the current target environment node, where the target-category audio is audio that can remind the user to pay attention to the road environment;

[0022] Step S520: Play audio of the target category to the user.

[0023] Optionally, step S510 includes:

[0024] Collect environmental sounds from the current target environment node;

[0025] Extract audio of target categories from ambient sounds;

[0026] Step S520 includes: playing audio of the target category simultaneously while the earphone is playing media audio.

[0027] In a second aspect, an embodiment of the present invention discloses a road condition information prompting device based on machine learning, comprising:

[0028] A static posture acquisition module is used to obtain static posture parameters of the headset when it is worn. The static posture parameters include a first static parameter and a second static parameter. The first static parameter is a gravity sensor posture parameter corresponding to the user looking forward when wearing the headset, and the second static parameter is a gravity sensor posture parameter corresponding to the user looking down when wearing the headset.

[0029] The dynamic parameter acquisition module is used to obtain dynamic parameters by using the gravity sensor to collect the dynamic posture of the user's head when the user moves along a preset route. The dynamic parameters are the parameters of the gravity sensor relative to the static posture parameters when the user wears the headset. The preset route is the route the user moves from the starting point to the destination.

[0030] The posture learning module is used to iteratively learn the user's behavioral posture during the process of moving along the preset route based on dynamic parameters to record the user's movement habits along the preset route;

[0031] A node marking module for being reminded is used to mark nodes to be reminded in a preset route. The nodes to be reminded include nodes whose corresponding dynamic parameters change more than a threshold relative to the static posture parameters;

[0032] The prompt module is used to send a first prompt message to the user when the user moves to the node to be reminded on the preset route, so as to remind the user to pay attention to the road condition information.

[0033] Optionally, in the prompt module, the first prompt information is a specific audio prompt tone.

[0034] Optionally, it also includes:

[0035] An ambient sound collection module is used to collect ambient sounds through a microphone when the user moves along a preset route;

[0036] The target recognition module is used to recognize environmental sounds and execute the target marking module when the target category sound is recognized in the environmental sounds;

[0037] A target marking module is used to mark a target environment node in a preset route, where the target environment node is a node in the preset route where sounds of the target category exist;

[0038] The prompt module also includes:

[0039] The second prompt unit is used to send a second prompt message to the user when the user moves to the target environment node along the preset route, so as to remind the user to pay attention to the current environment.

[0040] Optionally, the prompt module includes:

[0041] An extraction unit, configured to extract audio of a target category from the current target environment node, where the audio of the target category is audio that can prompt the user to pay attention to the road environment;

[0042] The playback unit is used to play audio of the target category to the user.

[0043] Optionally, the extraction unit is used to collect environmental sounds from the current target environment node; extract audio of the target category from the environmental sounds;

[0044] The playback unit is used to simultaneously play audio of the target category while the earphone is playing media audio.

[0045] In a third aspect, an embodiment of the present invention discloses a computer-readable storage medium having a computer program stored thereon. The computer program stored in the storage medium is used to be executed to implement the method disclosed in the first aspect above.

[0046] In a fourth aspect, an embodiment of the present invention discloses a chip for an audio device, which has an integrated circuit thereon, and the integrated circuit is designed to implement the method disclosed in the first aspect above.

[0047] In a fifth aspect, an embodiment of the present invention discloses a headset, comprising:

[0048] A processor is used to implement the method disclosed in the first aspect above.

[0049] According to an embodiment of the present invention, a method, device and earphones for prompting road condition information based on machine learning are disclosed. When the earphones are worn, the static posture parameters of the gravity sensor are obtained. Then, when the user moves on a preset route, the dynamic posture of the user's head can be collected by the gravity sensor to obtain dynamic parameters. Thus, the user's behavioral posture can be iteratively learned through the change of the dynamic parameters relative to the static posture parameters. That is, the user's movement habits on the preset route can be recorded, and the nodes whose changes exceed the threshold are marked as nodes to be reminded. Therefore, when the user moves to the node to be reminded on the preset route, the first prompt information is sent to the user, thereby reminding the user to pay attention to the road condition information. During the entire movement process of the user, the road condition information can be prompted to the user without the user taking off the earphones, which will not affect the user experience of wearing the earphones. Subsequently, it is achieved that in the process of the user using the earphones, the road condition information is effectively prompted to the user while minimizing the user's experience of using the earphones.

[0050] Other beneficial effects of the present invention will be explained through the introduction of specific technical features and technical solutions in the specific implementation methods. Those skilled in the art should be able to understand the beneficial technical effects brought about by the introduction of these technical features and technical solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The embodiments of the present invention will be described below with reference to the accompanying drawings.

[0052] Figure 1 This is a flow chart of a method for providing road condition information based on machine learning disclosed in this embodiment;

[0053] Figure 2 This is a schematic diagram of an example of a user's commute route disclosed in this embodiment;

[0054] Figure 3 This is a schematic diagram of an example of a road condition for a user's commute to work disclosed in this embodiment;

[0055] Figure 4 This is a schematic diagram of another example of road conditions for a user's commute to work disclosed in this embodiment;

[0056] Figure 5 This is a schematic diagram of the structure of a road condition information prompt device based on machine learning disclosed in this embodiment. DETAILED DESCRIPTION

[0057] The present invention is described below based on the following embodiments, but the present invention is not limited to these embodiments. In the following detailed description of the present invention, some specific details are described in detail. In order to avoid obscuring the essence of the present invention, well-known methods, processes, procedures, and components are not described in detail.

[0058] Furthermore, persons of ordinary skill in the art will appreciate that the figures provided herein are for illustration purposes only and are not necessarily drawn to scale.

[0059] Unless the context clearly requires otherwise, throughout the specification and claims, the words "include," "comprising," and similar words should be construed in an inclusive sense rather than an exclusive or exhaustive sense; that is, in the sense of "including but not limited to."

[0060] In the description of the present invention, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood to indicate or imply relative importance. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0061] In order to effectively prompt the user with traffic information while minimizing the impact on the user's experience of using headphones, this embodiment discloses a traffic information prompting method based on machine learning. Figure 1 , is a flow chart of a method for prompting traffic information based on machine learning disclosed in this embodiment. The method for prompting traffic information based on machine learning includes: step S100, step S200, step S300, step S400 and step S500, wherein:

[0062] Step S100: Obtain static posture parameters when the headset is worn. In this embodiment, the static posture parameters include a first static parameter and a second static parameter, wherein the first static parameter is the gravity sensor posture parameter corresponding to the user's forward-looking state when wearing the headset, and the second static parameter is the gravity sensor posture parameter corresponding to the user's head-down state when wearing the headset. Specifically, headsets are generally equipped with a gravity sensor (G-sensor). The parameters displayed by the gravity sensor are different when the user is in different states. In this embodiment, by obtaining the posture parameters of the gravity sensor when the user is looking forward and looking down, the baseline coordinate posture parameters can be used as the user's forward-looking and head-down posture parameters when using the mobile phone. When the user's posture changes, the user's posture can be compared with the static posture parameters to determine the user's changed posture. In a specific implementation process, a corresponding APP can be installed on a mobile terminal such as a mobile phone, and the user triggers the acquisition of the static posture parameters. After the headset and the mobile phone are connected, the user triggers the acquisition of the static posture parameters by operating the mobile phone after wearing the headset. It should be noted that, in a specific implementation process, for the same user, it is only necessary to obtain the static posture parameters once, that is, when executing subsequent steps, there is no need to obtain the static posture parameters again.

[0063] Step S200: When the user moves along a preset route, the dynamic posture of the user's head is collected by a gravity sensor to obtain dynamic parameters.

[0064] In this embodiment, the so-called dynamic parameters are the parameters of the gravity sensor relative to the static posture parameters when the user wears headphones. Specifically, when the user moves while wearing headphones, changes in movement angle and speed will cause changes in the parameters collected by the gravity sensor. Therefore, the dynamic posture of the user's head can be obtained to obtain dynamic parameters, such as angle, acceleration, etc.

[0065] In this embodiment, the so-called preset route is the route that the user moves from the initial point to the destination, specifically the route that the user often walks or rides, such as the route to school, the route to work, etc. Figure 2For the user working route example schematic diagram disclosed in the embodiment, the starting point is "residence", the destination is "office building", and the user will pass through motorway, overpass, residential area and the like in the user's advancing direction. For different road conditions, the user's posture will change. For example, the user will slow down at the intersection, and at this time, the parameters collected by the gravity sensor can reflect that the acceleration is negative. For another example, when there is a manhole cover on the road surface, the user will slow down and lower his head to observe the road surface, and at this time, the parameters collected by the gravity sensor can reflect that the acceleration is negative and the head posture is closer to the second static parameter when the user lowers his head. In the embodiment, when the user moves on the preset route, the movement habit of the user in the movement process on the preset route can be learned by collecting the head movement posture of the user, and the movement habit can represent the current road condition information, so that the user can be reminded of the road condition information when walking on the same preset route again.

[0066] In an optional embodiment, a preset time window (for example, a usual working time period and a usual leaving time period) of walking a preset route can be acquired. The earphone will only detect and train for a long time in the preset time window, so as to avoid false detection and influence on subsequent training accuracy caused by short-time walking of the user in a non-preset route. Specifically, the preset time window period can be set through the mobile phone after the earphone is connected to the mobile phone.

[0067] In step S300, the behavior posture of the user in the movement process on the preset route is learned according to the dynamic parameters, so as to record the movement habit of the user on the preset route.

[0068] In a specific embodiment, by comparison with the static posture parameters, the training samples of the user in different states in the whole process can be acquired, and the change of the head acceleration is added to train the training model and the detection model of the whole trip route.

[0069] In order to avoid false triggering of learning, in a specific implementation process, when it is detected that the user is at the starting point of the preset route and the speed of the user acquired in the preset time window is in a preset first average speed interval (for example, walking speed) or in a preset second average speed interval (for example, riding speed), the earphone will start to record the head posture data, walking acceleration and relatively maintained time of the user; or when it is detected that the user is at the starting point of the preset route and continuously maintains a relatively fixed speed after a period of time, that is, the head acceleration of the user acquired by the gravity sensor tends to 0, the earphone will start to record the head posture data, walking acceleration and relatively maintained time of the user.

[0070] In this embodiment, the head speed and change time obtained from the user's initial action can be used to calculate the user's habitual starting speed. Combined with the subsequent time the user continues walking, the speed and head posture characteristics used by the user in a relatively straight section can be calculated based on the horizontal coordinate changes fed back by the sensor. In this way, the user's behavioral posture during the process of moving along the preset route can be iteratively learned, thereby recording the user's movement habits along the preset route.

[0071] In a specific embodiment, whenever a user arrives at a preset time window, the user's behavioral posture when moving on a preset route can be iteratively learned, so that the behavioral posture recorded after the iteration is closer to the user's actual posture, thereby more accurately reflecting the user's movement habits.

[0072] Step S400, marking the nodes to be reminded in the preset route. In this embodiment, the so-called nodes to be reminded include nodes whose corresponding dynamic parameters change by more than a threshold value relative to the static posture parameters. Specifically, when the headset detects that the user changes in a relatively constant state of motion, it indicates that there are road conditions at the node that require attention. At this time, the node can be marked as a node to be reminded. In the specific implementation process, the change of the user's relatively constant state of motion can be reflected by acceleration, for example, when changing from a uniform speed to acceleration / deceleration, there is a change in acceleration; of course, it can also be reflected by angle, for example, when the user arrives at an intersection, he will turn his head to look at the road conditions. In a specific embodiment, the threshold value can be determined based on actual experience.

[0073] Step S500, when the user moves to the node to be reminded on the preset route, a first prompt message is sent to the user to remind the user to pay attention to the road condition information. Specifically, the first prompt message is a specific audio prompt tone. The first prompt message can be a prompt tone preset by the system, or a prompt tone set by the user, or an ambient sound collected in real time. When the first prompt tone is an ambient sound, only some key sounds (such as engine sounds, construction sounds) can be extracted, and the ambient sound can be further subjected to noise reduction processing to highlight the key sounds. In an optional embodiment, when the first prompt message is sent, the music or call played by the media can be kept uninterrupted, thereby avoiding interrupting the user's use of headphones.

[0074] After marking the node to be reminded, in one embodiment, when the user moves to the node to be reminded, a first prompt message is sent to the user; in another embodiment, when the user moves to the node to be reminded and the user's posture changes relative to the previous posture (that is, the user does not move according to the previous habit), the first prompt message is sent to the user. In other words, when the user moves according to the previous habit, the first prompt message is not sent, which can further reduce the user's headphone experience.

[0075] It should be noted that the step S200, the step S300 and the step S400 are a process of continuous iteration, that is, before the current execution of the step S500, the step S200, the step S300 and the step S400 can be iterated, and when the current execution of the step S500 is performed, for the node to be reminded, the last iteration result can be used, or the result updated this time can be used.

[0076] For the convenience of those skilled in the art, please refer to Figure 2 , Figure 2 The route from home to the company of the example user is shown, and the road condition environment is complex, which will have straight road sections, overpass road sections, T-shaped road sections, non-motor vehicle road sections and the like.

[0077] With reference to the above route, in an embodiment, the user starts from the habitual starting point (for example, the user's residence), and the earphone obtains a habitual starting speed of the user, and in the next straight road section with familiar road conditions, the user will maintain a relatively stable walking speed and habit.

[0078] At the first T-shaped intersection (the rotation speed collected by the gravity sensor can be used to determine the user's turning), the intersection is used to change before the vehicle condition intersection, the user will habitually make a large turning, and may make a certain observation, at this time, the motion state of the user will change relatively obviously compared with the previous straight road section.

[0079] At the second intersection, since the road condition of the intersection is relatively complex, if the user is in the state of walking and looking down at the mobile phone before, at this time, he / she will probably look up to observe the road condition, and the walking speed will also change relatively obviously. And at this time, since the bridge hole needs to be passed, different types of vehicles and pedestrians will meet, so the road section will become crowded, and the walking speed will also change relatively obviously.

[0080] The earphone obtains the gravity sensor parameters (such as acceleration, rotation speed, etc.) fed back by the gravity sensors of the left and right ears, calculates the advancing direction of the user, and can obtain the head posture data and the advancing direction data of the user through the changes of the gravity sensor feedback, and generate a road condition reminding training model. Specifically, through the gravity sensor parameters fed back by the gravity sensors of the left and right ears, the angle can be calculated, and the physical movement of the head shaking can be recognized.

[0081] Please refer to Figure 3 , for a road condition example schematic diagram in a user's working route disclosed by the embodiment, Figure 3As shown, the user reaches a T-shaped intersection, and the road condition at the intersection changes compared to the previous straight road section. According to the user's usual walking head posture and walking route, the user will habitually make a large turn to cross the road through the sidewalk. When the user reaches the T-shaped intersection, or the user does not observe the road information correctly, looks down at the mobile phone, etc., the first prompt information can be sent to remind the user to make certain observations. At this time, the user's movement condition will change more obviously compared to the previous straight road section.

[0082] In an optional embodiment, before step S400, the method further includes: step S210, collecting environmental sound through a microphone when the user moves along the preset route; step S220, identifying the environmental sound, and when a sound of a target category is identified in the environmental sound, performing step S410; and step S410, marking a target environmental node in the preset route, the target environmental node being a node in the preset route where the sound of the target category exists. Step S500 further includes: when the user moves to the target environmental node along the preset route, sending second prompt information to the user to remind the user to pay attention to the current environment. Specifically, step S210 and step S200 can be interleaved or performed synchronously, that is, dynamic parameters and environmental sound can be collected synchronously when the user moves along the preset route. Step S410 and step S400 can be interleaved or performed synchronously, that is, the node can be marked when it needs to be marked. In this embodiment, the target category sound refers to the sound corresponding to an event that is dangerous to the user, such as engine sound or construction sound. In the specific implementation process, road condition information can be learned by simultaneously collecting dynamic parameters and environmental sound. For example, when the user's posture changes while collecting construction sound, it indicates that the node has construction and is located on the preset route, so the node can be marked for reminding and target environmental node.

[0083] Please refer to Figure 4 , a user's work route is disclosed in another road condition example, Figure 4 As shown, a non-motor vehicle road section is reached, and the road condition at the intersection changes compared to the previous road section. According to the user's usual walking head posture and walking route, it is recorded that the user will pass around the stone pier (for example, sideways or small-angle turning) when crossing the zebra crossing. The user can be reminded to pay attention to the stone pier when crossing the zebra crossing to avoid bumping. At the same time, the construction sound of the manhole cover can be learned by the machine, so at this node, the user can also be reminded to pay attention to whether the manhole cover is under construction to avoid stepping on the empty space. In addition, this road section belongs to a non-motor vehicle road section, and attention should be paid to the coming and going bicycles, electric vehicles, etc. At this time, the user's movement condition will change more obviously compared to the previous straight road section.

[0084] In an optional embodiment, in step S500, sending a second prompt message to the user includes: step S510, extracting audio of the target category from the current target environment node, the audio of the target category is audio that can remind the user to pay attention to the road environment; step S520, playing the audio of the target category to the user. The target category refers to the event category that may pose a safety hazard to the user's movement, such as the sound of motor vehicles, electric vehicles, construction, etc. In the specific implementation process, the audio of the target category can be a system preset sound, or it can be a real-time collection and extraction of key sounds, such as engine sounds, horn sounds, bicycle horns, electric vehicle horns, construction sounds, etc.

[0085] Of course, in the specific implementation process, when implementing the method of collecting and extracting key sounds, noise can be filtered out to make the key sounds purer.

[0086] In an optional embodiment, step S510 includes: collecting ambient sound from the current target environment node; extracting audio of the target category from the ambient sound; and step S520 includes: playing the audio of the target category simultaneously with the headphone playing media audio. Specifically, the energy of the key sound can be relatively low, just enough to provide a reminder and avoid affecting the user's listening to music or making calls.

[0087] This embodiment also discloses a road condition information prompting device based on machine learning, please refer to Figure 5 , is a schematic diagram of the structure of a road condition information prompting device based on machine learning disclosed in this embodiment. The road condition information prompting device based on machine learning includes: a static posture acquisition module 100, a dynamic parameter acquisition module 200, a posture learning module 300, a node marking module to be reminded 400, and a prompting module 500, wherein:

[0088] The static posture acquisition module 100 is used to obtain the static posture parameters of the headset when being worn. The static posture parameters include a first static parameter and a second static parameter. The first static parameter is the gravity sensor posture parameter corresponding to the user looking forward when wearing the headset, and the second static parameter is the gravity sensor posture parameter corresponding to the user looking down when wearing the headset.

[0089] The dynamic parameter acquisition module 200 is used to obtain dynamic parameters by using a gravity sensor to collect the dynamic posture of the user's head when the user moves along a preset route. The dynamic parameters are the parameters of the gravity sensor relative to the static posture parameters when the user wears headphones. The preset route is the route the user moves from the starting point to the destination.

[0090] The posture learning module 300 is used to iteratively learn the user's behavioral posture during the process of moving along the preset route based on dynamic parameters, so as to record the user's movement habits along the preset route;

[0091] The node marking module 400 is used to mark nodes to be reminded in the preset route. The nodes to be reminded include nodes whose corresponding dynamic parameters relative to the static posture parameters change by more than a threshold value.

[0092] The prompt module 500 is used to send a first prompt message to the user when the user moves to a node to be reminded along a preset route, so as to remind the user to pay attention to road condition information.

[0093] In an optional embodiment, in the prompt module 500, the first prompt information is a specific audio prompt tone.

[0094] In an optional embodiment, the method further includes:

[0095] An ambient sound collection module is used to collect ambient sounds through a microphone when the user moves along a preset route;

[0096] The target recognition module is used to recognize environmental sounds and execute the target marking module when the target category sound is recognized in the environmental sounds;

[0097] A target marking module is used to mark a target environment node in a preset route, where the target environment node is a node in the preset route where sounds of the target category exist;

[0098] The prompt module 500 further includes:

[0099] The second prompt unit is used to send a second prompt message to the user when the user moves to the target environment node along the preset route, so as to remind the user to pay attention to the current environment.

[0100] In an optional embodiment, the prompt module 500 includes:

[0101] An extraction unit, configured to extract audio of a target category from the current target environment node, where the audio of the target category is audio that can prompt the user to pay attention to the road environment;

[0102] The playback unit is used to play audio of the target category to the user.

[0103] In an optional embodiment, the extraction unit is configured to collect environmental sounds from the current target environment node; extract audio of the target category from the environmental sounds;

[0104] The playback unit is used to simultaneously play audio of the target category while the earphone is playing media audio.

[0105] This embodiment further discloses a computer-readable storage medium, such as a chip, an optical disc, etc., on which a computer program is stored. The computer program stored in the storage medium is used to be executed to implement the method disclosed in the above embodiment.

[0106] This embodiment further discloses a chip of an audio device, which has an integrated circuit thereon. The integrated circuit is designed to implement the method disclosed in the above embodiment.

[0107] This embodiment further discloses a headset, which may be a wired headset or a wireless headset. The headset includes: a processor for implementing the method disclosed in the above embodiment.

[0108] According to an embodiment of the present invention, a method, device and earphones for prompting road condition information based on machine learning are disclosed. When the earphones are worn, the static posture parameters of the gravity sensor are obtained. Then, when the user moves on a preset route, the dynamic posture of the user's head can be collected by the gravity sensor to obtain dynamic parameters. Thus, the user's behavioral posture can be iteratively learned through the change of the dynamic parameters relative to the static posture parameters. That is, the user's movement habits on the preset route can be recorded, and the nodes whose changes exceed the threshold are marked as nodes to be reminded. Therefore, when the user moves to the node to be reminded on the preset route, the first prompt information is sent to the user, thereby reminding the user to pay attention to the road condition information. During the entire movement process of the user, the road condition information can be prompted to the user without the user taking off the earphones, which will not affect the user experience of wearing the earphones. Subsequently, it is achieved that in the process of the user using the earphones, the road condition information is effectively prompted to the user while minimizing the user's experience of using the earphones.

[0109] As an optional solution, the headset uses gravity sensors on both ears to obtain location information, along with acceleration data from the user to record walking routes and habits. This solution also uses specific ambient sounds to generate a training model. This model then provides road condition alerts and amplifies ambient sounds for different locations in different environments. This allows users to obtain road condition information without affecting their multimedia experience, such as listening to music, thereby improving the safety of walking for headphone wearers. Compared to traditional solutions that only amplify specific ambient sounds, this solution uses headphone environmental recognition models trained on the user's head posture and movement speed to amplify key ambient sound sources along specific road sections, allowing users to maintain their attention to road conditions without affecting their multimedia experience.

[0110] Moreover, since the headphone environment recognition model is obtained through training of the user's head posture and movement speed, it can continuously iterate and learn while the user moves along the preset route, amplifying the key environmental sound sources of specific sections of the road, allowing the user to still pay attention to the road conditions without affecting the multimedia experience such as listening to music.

[0111] It should be noted that the computer-readable storage medium in the embodiments of the present disclosure is not limited to the above-mentioned embodiments, for example, it can also be an electric, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or instrument, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to, an electric connection with one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiments of the present disclosure, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or instrument.

[0112] Those skilled in the art can understand that the above-mentioned preferred embodiments can be freely combined and superimposed without conflict. Among them, the flowcharts and block diagrams in the drawings illustrate the possible implementation architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram can represent a module, program segment, or part of code containing one or more executable instructions for implementing the specified logic function. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different order than that noted in the drawings, for example, two consecutive blocks can actually be executed substantially in parallel, and sometimes they can be executed in reverse order, depending on the function involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions. The numbering of the steps herein is only for the convenience of description and reference, and does not limit the front and back order, and the specific execution order is determined by the technology itself, and those skilled in the art can determine various allowed and reasonable orders according to the technology itself.

[0113] It should be noted that the step numbering (letter or number) is used in the present application to refer to certain specific method steps, only for the purpose of convenience and brevity, and by no means to limit the order of the method steps by letters or numbers. Those skilled in the art can understand that the order of the related method steps should be determined by the technology itself, and should not be improperly limited by the existence of step numbering, and those skilled in the art can determine various allowed and reasonable step orders according to the technology itself.

[0114] Those skilled in the art will appreciate that, provided there is no conflict, the above preferred solutions can be freely combined and superimposed.

[0115] It should be understood that the above-mentioned embodiments are merely illustrative and non-restrictive. Without departing from the basic principles of the present invention, various obvious or equivalent modifications or substitutions that can be made by those skilled in the art to the above-mentioned details will be included in the scope of the claims of the present invention.

Claims

1. A method for prompting road condition information based on machine learning, characterized in that: include: Step S100, obtaining static posture parameters of the headset when being worn, the static posture parameters including a first static parameter and a second static parameter, wherein the first static parameter is a gravity sensor posture parameter corresponding to a state in which the user is looking forward while wearing the headset, and the second static parameter is a gravity sensor posture parameter corresponding to a state in which the user is looking down while wearing the headset; Step S200: When a user moves along a preset route and a user speed acquired within a preset time window is within a preset first average speed interval or a second average speed interval, a gravity sensor is used to collect a dynamic head posture of the user to obtain a dynamic parameter, where the dynamic parameter is a parameter of the gravity sensor relative to the static posture parameter when the user wears headphones. The preset route is the route the user moves from an initial point to a destination. The first average speed interval corresponds to a walking speed, and the second average speed interval corresponds to a cycling speed. The dynamic parameter includes a parameter where acceleration is negative. Step S300, using the dynamic parameters as training samples of the user in different states, iteratively learning the user's behavior and posture during movement along the preset route through the training samples, so as to record the user's movement habits along the preset route; Step S400, marking nodes to be reminded in the preset route, wherein the nodes to be reminded include nodes whose corresponding dynamic parameters relative to the static posture parameters vary by more than a threshold value; Step S500: When the user moves to the node to be reminded along the preset route, a first reminder message is sent to the user to remind the user to pay attention to road condition information.

2. The method for prompting road condition information based on machine learning according to claim 1, characterized in that: In the step S500, the first prompt information is a specific audio prompt tone.

3. The method for prompting road condition information based on machine learning according to claim 1 or 2, characterized in that: Before step S400, the method further includes: Step S210, when the user moves along the preset route, the microphone collects ambient sound; Step S220, identifying the ambient sound. When the target sound is detected in the ambient sound, executing step S410; Step S410, marking a target environment node in the preset route, wherein the target environment node is a node in the preset route where a sound of a target category exists; The step S500 further includes: When the user moves to the target environment node along the preset route, a second prompt message is sent to the user to remind the user to pay attention to the current environment.

4. The method for prompting road condition information based on machine learning according to claim 3, characterized in that: In step S500, the sending of the second prompt information to the user includes: Step S510: extracting target-category audio from the current target environment node, where the target-category audio is audio that can prompt the user to pay attention to the road environment; Step S520: Play audio of the target category to the user.

5. The method for prompting road condition information based on machine learning according to claim 4, characterized in that: The step S510 includes: Collect environmental sounds from the current target environment node; extracting audio of a target category from the environmental sounds; The step S520 includes: playing the audio of the target category simultaneously while the earphone is playing the media audio.

6. A road condition information prompting device based on machine learning, characterized in that: include: A static posture acquisition module (100) is used to acquire static posture parameters when the headset is worn, the static posture parameters comprising a first static parameter and a second static parameter, wherein the first static parameter is a gravity sensor posture parameter corresponding to a state in which the user is looking forward when wearing the headset, and the second static parameter is a gravity sensor posture parameter corresponding to a state in which the user is looking down when wearing the headset; A dynamic parameter acquisition module (200) is used to acquire dynamic parameters by collecting the dynamic posture of the user's head through a gravity sensor when the user moves along a preset route and the user speed acquired within a preset time window is within a preset first average speed interval or a second average speed interval, wherein the dynamic parameters are parameters of the gravity sensor relative to the static posture parameters when the user wears headphones; the preset route is the route the user moves from an initial point to a destination; the first average speed interval corresponds to a walking speed, and the second average speed interval corresponds to a cycling speed; the dynamic parameters include: parameters with negative acceleration; A posture learning module (300) is used to use the dynamic parameters as training samples of the user in different states, and iteratively learn the user's behavioral posture during the process of moving along the preset route through the training samples, so as to record the user's movement habits along the preset route; A node marking module (400) for marking nodes to be reminded in the preset route, wherein the nodes to be reminded include nodes whose corresponding dynamic parameters relative to the static posture parameters vary by more than a threshold value; The prompt module (500) is used to send a first prompt message to the user when the user moves to the node to be reminded along the preset route, so as to remind the user to pay attention to the road condition information.

7. The road condition information prompting device based on machine learning according to claim 6, characterized in that: In the prompt module (500), the first prompt information is a specific audio prompt tone.

8. The machine learning-based road condition information prompting device according to claim 6 or 7, characterized in that: Also includes: An ambient sound collection module is used to collect ambient sounds through a microphone when the user moves along a preset route; a target recognition module, configured to recognize the ambient sound and execute the target marking module when a target-class sound is recognized in the ambient sound; a target marking module, configured to mark a target environment node in the preset route, wherein the target environment node is a node in the preset route where a sound of a target category exists; The prompt module (500) further includes: The second prompting unit is used to send a second prompting message to the user when the user moves to the target environment node along the preset route, so as to remind the user to pay attention to the current environment.

9. The machine learning-based road condition information prompting device according to claim 8, characterized in that: The prompt module (500) includes: An extraction unit, configured to extract audio of a target category from a current target environment node, wherein the audio of the target category is audio that can prompt a user to pay attention to the road environment; The playback unit is used to play audio of the target category to the user.

10. The device for prompting road condition information based on machine learning according to claim 9, characterized in that: The extraction unit is used to collect environmental sounds from the current target environment node; extract the target category of audio from the environmental sounds; The playing unit is configured to simultaneously play the audio of the target category while the earphone is playing the media audio.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program stored in the storage medium is used to be executed to implement the method according to any one of claims 1 to 5.

12. A chip for an audio device having an integrated circuit thereon, characterized in that: The integrated circuit is designed to implement the method according to any one of claims 1-5.

13. A headset, characterized in that: include: A processor, configured to implement the method according to any one of claims 1 to 5.

Images