ByteDance Patente

The present disclosure involves methods (200), apparatus, and systems (100, 300) for multi-layer noise generation in differential privacy (DP). This can include identifying (202) an SQL query; performing (206) an SQL rewrite of the SQL query by replacing at least one process with a DP process; executing (208) the SQL query, and while executing the SQL query injecting noise during the DP process by: generating (210A) random noise; generating (210B, 210C) fixed noise; combining (210D), as combined noise, the random noise and the fixed noise in a weighted sum; and injecting the combined noise during the DP process.

The present disclosure involves methods (200), apparatus, and systems (100, 700) for selection operations to perform a differential privacy (DP) rewrite. This can include: identifying (202) a SQL query; identifying (204) one or more source tables against which the SQL query will be processed and a classification of each source table; generating (206) a parse tree based on the SQL query, the parse tree comprising the one or more source tables, an output table, and at least one operation; searching (210) the parse tree for operations to re-write by: selecting a current branch to search; performing, as an analyzed operation, an analysis of each operation in the current branch starting at the output table and for each operation: determining (212) whether the analyzed operation is an aggregation process that transforms a user-level detailed table (ULDT) into an aggregate table.

The present disclosure involves methods (200, 300), apparatus, and systems (100, 400) for dynamically determining sensitivity in differential privacy. This can include: identifying a query for aggregate information from a source, wherein the aggregate information is associated with a plurality of users; grouping each piece of information in the source based on an associated user for that piece of information; determining, for each associated user, a primary aggregation, wherein the primary aggregation is a vector representing results to the query sorted by each associated user's contribution to the source; calculating, as a calculated user norm, a user norm for each user's primary aggregation; identifying a subset one or more users whose primary aggregation is greater than the sensitivity associated with the source; and truncating subset one or more users' primary aggregation by a ratio of their associated calculated user norm and the sensitivity.

The present disclosure provides a method of waking up an application, an electronic device, and a storage medium. The method includes: obtaining target wake-up information (S201); determining, among a plurality of agent objects, a target agent object corresponding to the target wake-up information (S202); and waking up the target agent object (S203).

An interaction control method, a device, and a storage medium are provided. The method includes: acquiring target finger motion information of a user, where the target finger motion information includes a motion speed of a finger and/or a displacement of the finger (S110); determining target state change information of an object to be controlled according to the target finger motion information (S120); and adjusting a state of the object to be controlled according to the target state change information, where different finger motion information corresponds to different state change information (S130). In this way, energy consumption of the device during interaction is reduced, user operations are simplified, and precise control of the object to be controlled is achieved.

The present application provides an earphone and an earphone system, where the earphone includes a sound emitting part, a main body part and an ear hanging part connecting the sound emitting part and the main body part; wherein the sound emitting part includes a first housing, a camera apparatus and a control module, where the first housing is provided with a visual field hole, and the camera apparatus is arranged at the visual field hole,; the control module includes a main board, an image processing unit and a main control unit, where the main board includes a first board body, a second board body and a flexible connecting board which are connected in sequence; the image processing unit is arranged on the first board body; and the main control unit is arranged on the second board body.

Embodiments of the present disclosure provide an interaction control method, an interaction control device, an electronic apparatus and a storage medium, which relate to the technical field of headsets. The method includes: acquiring an interaction state with a terminal apparatus; receiving a voice signal through a microphone; in the case where a bone sound signal is received through a bone sound sensor, determining whether the voice signal is emitted by a wearer at least according to the bone sound signal; and in the case where the voice signal is emitted by the wearer, performing interaction processing with the terminal apparatus based on the interaction state according to the voice signal. Interaction with the terminal apparatus is performed only when it is determined that the voice signal is emitted by the wearer by combining the bone sound signal, thereby solving the problem of a high probability of interaction errors caused by environmental noise.

Embodiments of the present disclosure provide an image processing method, a wearable device (11), an electronic device, and a storage medium. The method includes: determining a target resolution in response to a shooting instruction; and controlling a camera to shoot a first image (a1) with the target resolution for a current environment, where the target resolution is a first resolution or a second resolution, and the first resolution is greater than the second resolution. In the present disclosure, the wearable device (11) may be controlled to shoot images with different resolutions, so that the wearable device (11) may be applied to various scenarios.

A method, device and storage medium for content delivery are provided. In the method, a request to deliver a target content to a target object is received. For a content delivery strategy of a plurality of content delivery strategies, at least one group of observed samples are determined based on target features of the target content and the target object, an observed sample comprising reference features of a reference content and a reference object, an indication of whether the content delivery strategy is applied to deliver the reference content and an object response. An effect metric is determined for the content delivery strategy based on the at least one group of observed samples. A target content delivery strategy for delivering the target content to the target object is selected from the plurality of content delivery strategies based on respective effect metrics for the plurality of content delivery strategies.

According to embodiments of the disclosure, a method, an apparatus, a device, a medium, and a product for display processing are provided. The method includes: determining a plurality of rendering layers to be rendered on a display screen, each rendering layer corresponding to at least a partial interface area in an application interface to be displayed; in response to a number of the plurality of rendering layers being greater than a threshold number, determining a respective priority of each of the plurality of rendering layers, the threshold number indicating an upper limit of a number of rendering layers that may be rendered within a predetermined period; performing rendering on the plurality of rendering layers within the predetermined period based on the respective priority of each of the plurality of rendering layers to obtain a rendering result; and presenting the rendering result on the display screen.

Embodiments of the present disclosure relate to a method, an apparatus, a device, and a computer-readable storage medium for generating skeleton data. The method (200) proposed herein includes: sampling (210), based on shape data of an object, a plurality of surface points of the object; generating (220) a set of shape units by encoding the plurality of surface points; providing (230) the set of shape units to a model to generate a set of skeleton units; and generating (240) skeleton data of the object based on the set of skeleton units, the skeleton data indicating a plurality of joint points of the object and a connection relationship between the plurality of joint points.

An optical system and a display apparatus are provided. The optical system includes an optical waveguide (100) and a Fresnel lens (200). The optical waveguide (100) includes a waveguide substrate (110), a coupling-in structure (120) and a coupling-out structure (130), the coupling-in structure (120) and the coupling-out structure (130) are provided on the waveguide substrate (110), the optical waveguide (100) is configured such that at least part of light coupled into the optical waveguide (100) via the coupling-in structure (120) exits via the coupling-out structure (130) after undergoing total internal reflection propagation in the optical waveguide (100). The Fresnel lens (200) is located on a light-exiting side of the optical waveguide (100). The coupling-out structure (130) includes a grating (131,132), the Fresnel lens (200) includes Fresnel teeth (201), and the Fresnel teeth (201) are configured to adjust an exit angle of at least part of light passing through the Fresnel lens (200). In a direction perpendicular to the waveguide substrate (110), the coupling-out structure (130) and the Fresnel teeth (201) overlap. Therefore, the Fresnel lens (200) can be used for refractive correction, so that the optical system and the display apparatus are thinner, and at the same time, users with refractive errors can see the display screen clearly.

Embodiments of the present disclosure proposes a headset-based text recognition method, apparatus, headset, storage medium and program product, and relates to the field of headset technology. The method is applied to a processing unit, and includes: when it is detected that a text recognition mode is triggered, sending a shooting instruction to a camera unit, so that the camera unit can acquire at least one image to be recognized of the recognized object according to the shooting instruction; receiving the at least one image to be recognized sent by the camera unit; inputting the at least one image to be recognized into a preset text recognition model, so that the preset text recognition model can recognize the image to be recognized, so as to obtain text content; and outputting the text content.

The present disclosure relates to a scheduling method and apparatus, and a computer-readable storage medium, and relates to the field of computer technologies. The scheduling method includes: determining, according to a target device of a first operation request, a device priority corresponding to the first operation request; and determining, according to the device priority and a currently idle hardware resource, whether to issue the first operation request to the target device for processing.

The present disclosure relates to an interaction method and apparatus, an electronic device, and a storage medium. The method includes: displaying a video playback window within an extended reality space, the video playback window being configured to display image content of a local area in a VR video image; and in response to an adjustment operation for a VR video picture, keeping the video playback window fixed in position within the extended reality space, and adjusting the local area of the VR video image displayed in the video playback window, a picture displayed in the video playback window after the adjustment being not exactly the same in content as a picture displayed in the video playback window before the adjustment. A method is provided which allows a user to adjust a viewing angle of a VR video according to his/her own needs to satisfy personalized viewing needs of the user.

The present application discloses a media data recommendation method and apparatus, a device, a storage medium, and a program product, which relate to the technical field of data recommendation. The method includes: acquiring a first interactive instruction for a first multimedia application, where the first interactive instruction is used to trigger display of a first page; and displaying the first page, where first recommendation information and a source of the first recommendation information are displayed in the first page, the first recommendation information is information associated with an associated object of the first multimedia application in a second multimedia application, and the associated object is used to represent an object having an association relationship with a current object of the first multimedia application. The present disclosure may solve the problem of media data recommendation.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using differential privacy device bounded contributions. One of the methods includes maintaining (202), for each time period from a set of time periods, action data for a device; determining whether a current time period satisfies a time period criterion; selecting, using a result of the determination whether the time period criteria is satisfied and from a set of two or more bound generation processes, a bound generation process for the device; computing (212), using the bound generation process, a bound on an amount of data from the action data and for the device to transmit to a downstream system; selecting (214), using the action data for the device, a responsive data set with a size defined by the bound; and transmitting (220), to the downstream system, the responsive data set with the size defined by the bound.

Embodiments of the present disclosure provide a content generation method and apparatus, an electronic device and a storage medium. A target viewing angle is determined in response to a user operation, where the target viewing angle is configured to represent an observation viewing angle in a target state; an image medium is acquired, and a content generation model is invoked to process the image medium based on the target viewing angle to generate viewing angle effect content, where the viewing angle effect content has an effect video frame, and the effect video frame is configured to display a visual effect of observing the image medium based on the target viewing angle. A target viewing angle representing an observation viewing angle of a target observation object is determined, and an image generation model is invoked by using the target viewing angle as input to process an image medium to generate viewing angle effect content, thereby realizing video generation based on the viewing angle dimension, and improving the quality of the generated video.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for finding splits in gradient boosting systems. One example method includes identifying, based on multi-party computation (MPC), a first split among possible first splits in a first layer of a decision tree. Identifying the first split comprises determining a first derivative and a second derivative for each sample in a dataset of n samples and m features; generating a sorting array indicating a sorting order of at least one feature; identifying the first split based on first derivatives, second derivatives and the sorting array; and generating an index vector that indicates whether each sample belongs to a first or a second sub-dataset. The method further includes identifying a second split among possible second splits in a second layer of the decision tree based on the first derivatives, the second derivatives, the sorting array, and the index vector.

The embodiments of the present disclosure provides a speech playing method, an electronic device, and a storage medium, and relate to the field of computer technologies, where the method includes: obtaining first measurement data acquired by a sensor in a case where a user wears a headset; determining a behavior event of the user according to the first measurement data; inputting the behavior event into a pre-trained speech model for processing to obtain inquiry speech information; and playing the inquiry speech information. The present disclosure enables the headset to play inquiry speech information according to the first measurement data acquired by the sensor provided on the headset, and then enables active inquiry of the headset, without the need for the user to manually operate the headset or control the headset by operating a mobile terminal, thereby enabling improvement in control efficiency of the headset and enabling the headset to have an intelligent function of active inquiry.

The disclosure provides a headphone. The headphone includes: a sound generating part, configured to be worn on a front side of an ear, and including a speaker unit; a main body part, configured to be worn on a rear side of the ear; an ear clip part, opposite ends of which being respectively connected to the sound generating part and the main body part, and the ear clip part being configured to switch between a first form and a second form; a form detecting apparatus, provided on at least one of the sound generating part and the main body part, and configured to detect a wearing form of the ear clip part; and a control apparatus, configured to control an audio output state of the speaker unit according to a detection result of the form detecting apparatus.

The present disclosure provides an earphone, which relates to the technical field of wearable devices. The earphone includes a main body part and a sounding part, where the main body part straddles a helix of an ear in the wearing state, and the sounding part is located at one side of the main body part facing a concha cavity of the ear; the sounding part is slidably arranged on the main body part so that the sounding part has a first state and a second state relative to the main body part, and a distance between the sounding part and an ear canal opening in the second state is smaller than a distance between the sounding part and the ear canal opening in the first state.

The present disclosure relates to the field of wearable device technology, and in particular to an earphone and an earphone system, wherein the earphone includes: a sound-emitting portion configured to be disposed on a front side of an ear; a body portion configured to be disposed on a rear side of the ear; and a transition connecting portion, two ends of the transition connecting portion being respectively connected with the sound-emitting portion and the body portion, wherein the sound-emitting portion is detachably connected with the transition connecting portion.

An audio file generation method, a device, and a medium are provided. The method includes: displaying an editing control on a playing page; displaying an audio editing page in response to an interactive operation on the editing control, where the audio editing page includes audio information of a preset audio file and a generation control; determining a reference audio segment in the preset audio file in response to a selection operation on the audio information; and generating a target audio segment according to the reference audio segment and determining a target audio file according to the target audio segment in response to an interactive operation on the generation control. The method can generate a target audio segment with a similar style according to the reference audio segment selected by the user, which can meet the user's music generation expectation, and reduce the difficulty in music creation and improving user experience.

Embodiments of the present disclosure provide a media content generation method, a device, and a medium. The method includes: displaying a material import control and a generation control on a first page; displaying a second page in response to an interactive operation on the material import control, the second page including a candidate material; determining a target material in response to a selection operation on the candidate material, and displaying the target material on the first page, where the target material includes at least one of a target picture set and a target video; and generating a target audio according to the target material in response to an interactive operation on the generation control, and generating target media content according to the target audio and the target material.

The disclosure relates to the technical field of electronic products, in particular to an ear-clip earphone. The ear-clip earphone includes a speaker unit, a main body part and an ear clip part, where the speaker unit is adapted to be arranged on a front side of an ear, and the main body part is adapted to be arranged on a back side of the ear; both ends of the ear clip part are respectively connected to the speaker unit and the main body part, and the ear clip part is configured to switch between a first form and a second form, where in the first form, the speaker unit and the main body part jointly clamp the ear, and in the second form, the speaker unit abuts against the tragus. According to the ear-clip earphone provided by the disclosure, the user can switch the ear clip part between the first form and the second form according to different requirements, thus enriching the forms and functions of the ear-clip earphone and being beneficial to expanding the use scenarios of the ear-clip earphone.

Embodiments of the disclosure relate to a method, an apparatus, a device, and a storage medium for application interaction. The method proposed herein includes: receiving an operation of a user logging in a target application; determining a target identification material from a set of identification materials at a terminal device, the target identification material being determined based on reference information associated with the user; and updating an image identification of the target application by using the target identification material.

A singing video generation method and apparatus, an electronic device, and a storage medium are provided. After target music is generated by a target application, a publishing page corresponding to the target music is displayed in the target application, where a video generation control for generating a singing video matching the target music is configured in the publishing page; in response to a first trigger operation on the video generation control, a first video generation page is displayed in the target application, where the first video generation page is configured to display at least one video material; and in response to a second trigger operation on the first video generation page, a target video material is determined and the singing video is generated based on the target video material, where a mouth action of a target object in the singing video matches music content of the target music.

The embodiments of the disclosure provide a method, an apparatus, a device and a storage medium for interaction. The method includes: acquiring, in response to receiving an interaction request from a user, interaction content comprising at least image content through a content acquisition device; performing object detection on the image content to obtain an object detection result indicating at least one object in the image content; determining, based on the object detection result, a target object classified as a point of interest from the at least one object and an interaction requirement for the target object; and performing a target operation related to the target object based on the interaction requirement.

The embodiments of the present disclosure provide an AI song editing method and apparatus, an electronic device and a storage medium. First lyrics of a first song are displayed after the first song and the corresponding first lyrics are generated, where the first song is audio data generated based on artificial intelligence technology; second lyrics are generated in response to a modification instruction for first text in the first lyrics, where the second lyrics include second text corresponding to a play timestamp of the first text; and a second song is generated based on the second text in response to generating the second lyrics, where the second song includes a target song segment and a non-target song segment, the target song segment is a song segment corresponding to the second text, and a melody of the non-target song segment is the same as a melody of a corresponding song segment in the first song.

Embodiments of the present disclosure provide a method and an apparatus for protecting data in an electronic device, an electronic device, and a program product. The method (200) includes: receiving (202) first data from a first application by a security service, where the security service blocks the first application from accessing a second application to protect data. The method further includes: receiving (204) second data from the second application by the security service. In addition, the method further includes: determining (206) third data for the electronic device by the security service based on the first data and the second data.

Embodiments of the present disclosure provide an application interaction method and apparatus, an electronic device, and a storage medium, wherein a media object is played in a video playback interface based on an information stream in a first application, where a target control is provided in the video playback interface, and the target control dynamically displays a corresponding control icon based on a video feature of the media object; switching to a secondary operation interface of the first application is performed in response to a trigger operation for the target control, where the secondary operation interface is configured at least for displaying a media player or switching to a second application, the second application being configured for acquiring background music or a video template of the media object.

A microphone control system includes: a first switch and an audio processor; wherein a first selection terminal of the first switch is configured to be connected to a first microphone; a second selection terminal of the first switch is configured to be connected to a second microphone; a common terminal of the first switch is connected to a first microphone interface of the audio processor; the first switch is configured to receive a first switching control signal, under control of which the first selection terminal is connected to the common terminal to send a first audio signal received by the first microphone to the audio processor; and the first switch is further configured to receive a second switching control signal, under control of which the second selection terminal is connected to the common terminal to send a second audio signal received by the second microphone to the audio processor.

A speech interaction device, apparatus, method, and medium are provided. The speech interaction device includes, a speech processor (110), a first communication module (120), at least one microphone (130) and a speech interaction system (140), where the speech interaction system (140) includes a second communication module (141). A first input terminal of the speech processor (110) is connected with the microphone (130), a first output terminal of the speech processor (110) is connected with the first communication module (120), and the speech processor (110) is configured to determine that the microphone (130) obtains a first speech signal, establish a connection with a cloud server through the first communication module (120), and send the first speech signal obtained by the microphone (130) to the cloud server. The speech interaction system (140) is connected with the cloud server through the second communication module (141), and the speech interaction system (140) is configured to receive a first response signal sent by the cloud server for speech interaction.

An optical system and a near-eye display apparatus are provided. The optical system includes: a lens structure; and a beam splitting film, a reflective polarizing film and a phase retardation film disposed on the lens structure, the reflective polarizing film and the phase retardation film being located at a same side of the beam splitting film, image light transmitted through the beam splitting film being configured to be reflected between the beam splitting film and the reflective polarizing film and exit from the reflective polarizing film, and the lens structure including at least one lens. The optical system further includes a light filtering structure located at a side of the reflective polarizing film that is away from the beam splitting film, the light filtering structure is configured such that, among light rays incident from the side away from the reflective polarizing film, an average transmittance of light rays having an incidence angle greater than 50 degrees is a first transmittance, an average transmittance of light rays having an incidence angle less than 30 degrees is a second transmittance, and a ratio of the first transmittance to the second transmittance is less than 1/3. By arranging the light filtering structure having different transmittances for light rays with different incidence angles, it's conductive to reduce the interference of ambient light.

Embodiments of the present disclosure provide a video processing method, device, storage medium, and program product. The method includes: acquiring (S201) optical flow information of a to-be-processed video; acquiring a first intermediate video frame with a target object removed and an initial mask image for any initial video frame of the to-be-processed video; determining (S203) a pixel value of at least one target pixel in a to-be-repaired area of the first intermediate video frame according to the optical flow information and adjacent video frames of the initial video frame, and filling the target pixel of the first intermediate video frame and a corresponding pixel of the initial mask image respectively to obtain a second intermediate video frame and an intermediate mask image; and invoking (S204) a preset video repair model to repair the second intermediate video frame according to the optical flow information, the second intermediate video frame, and the intermediate mask image. Preliminary repair is performed by combining the optical flow information and the adjacent video frames, so as to preserve richer original textures, reduce the area of the area to be repaired, and then repair by using the video repair model, thus improving the repair efficiency and the erasing quality of the target object.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining whether to generate sub-domain noise. One of the methods includes, for a message for a downstream system, determining whether to generate noise data that has the same sub-domain as a true value, the true value from a domain that has a plurality of different sub-domains including the sub-domain; using a result of the determination whether to generate noise data that has the same sub-domain as the true value, generating the message for the downstream system; and transmitting, to the downstream system, the message.

Embodiments of the present disclosure provides a method and an apparatus for generating a media work, a device and a medium. The method includes: displaying a creation page in response to an operation instruction for a media application, where the creation page displays a creation control; displaying a work generation page in response to an interaction instruction for the creation control; displaying a first media work on the work generation page in response to an input instruction of description information in the work generation page, where the description information includes prompt information and a material; displaying a work processing page in response to a publication instruction of publishing the first media work to the media application; and obtaining a second media work in response to a processing instruction for the first media work on the work processing page, where the second media work is used for being published to the media application.

The present disclosure involves methods, apparatus, and systems for processing comparison in secure multiparty computation. One example method includes, partitioning a first difference (x) between a first share of value a and a first share of value b into N sections, and determining whether a<b based on first indicators and second indicators corresponding to the N sections. The first indicators include a first indicator indicating a comparison result between xj and yj computed based on a Vector Oblivious Shift Evaluation (VOSE) protocol, where xj is a j<th> section of x, and yj is a j<th> section of N sections of a second difference (y) between a second secret share of b and a second secret share of a. The second indicators include a second indicator indicating whether xj is a most significant section among sections of the first difference where xj ≠ yj.

The present disclosure provides an image processing method, a non-transitory storage medium, and an electronic device. The image processing method includes: obtaining a frame group to be processed in a target video; and processing the frame group to be processed by a saliency prediction model to obtain a detection result of the target video frame. The saliency prediction model includes a feature encoder and a feature decoder. The processing the frame group to be processed by a saliency prediction model includes: inputting the frame group to be processed into the feature encoder to perform feature extraction, and outputting multi-layer feature data corresponding to the frame group to be processed; and inputting the multi-layer feature data into the feature decoder to perform feature fusion in a temporal dimension and feature fusion in a spatial dimension, and outputting a detection result of the target video frames.