Method and device for generating a favorites list for a smart television, smart television and computer readable storage medium

Abstract

The application relates to the technical field of smart home, and discloses a collection folder generation method for a smart television, which comprises the following steps: in response to a collection triggering operation of a user on target media content, a collection entry containing a unique identifier of the content is generated; multi-dimensional context information is associated to the collection entry to construct a smart collection unit; in response to a combination function setting instruction for the target media content, a composite operation protocol for triggering a preset function sequence in one key is generated, and the composite operation protocol is bound to the smart collection unit; and the smart collection unit is synchronized to a plurality of terminal devices associated with a current user account, so that the smart collection unit remains in a callable state on the plurality of terminal devices. In this way, a collection system capable of intelligent sensing, dynamic adaptation and cross-device continuity is provided for the user, and the individualization and convenience experience of user operation is improved. The application further discloses a collection folder generation device for a smart television, a smart television and a computer readable storage medium.

Description

Technical Field

[0001] This application relates to the field of smart home technology, such as a method, apparatus, smart TV, and computer-readable storage medium for generating favorites for a smart TV. Background Technology

[0002] With the rapid iteration of smart TV functions and the immense enrichment of content, users are no longer faced with a vast library of programs, but also system settings nested under multiple menus, unique features distributed across different applications, and combined operation scenarios involving multiple steps. Early TV favorites functions focused on managing channel and program lists, such as automatically categorizing and recommending frequently watched channels based on user viewing habits. This approach initially solved the efficiency problem of users finding target programs among numerous channels, but it essentially remained a static archive of single content items.

[0003] To address the issues of complex operation levels and deeply nested function entry points in smart TVs, existing technologies have further developed a function path-based bookmarking solution. This technology discloses a system capable of bookmarking specific sub-function paths. Users can save the specific operation paths of frequently used functions to a bookmarks folder, enabling one-click access to functions from the bookmarks interface and avoiding the tedious process of searching through multiple menus. By combining operation paths with a bookmarking mechanism, this solution effectively solves the problems of users needing to remember function locations and lengthy operation steps, thus improving the convenience of function access.

[0004] While related technologies have solved the problem of quickly locating and activating specific functions, their collection logic remains simplistic and static. They only record the static path of the function, and collected content typically exists in isolation on a single device, failing to achieve state synchronization and scene continuity across different user terminals. Therefore, how to build a collection system that enables intelligent cross-device synchronization and dynamic adaptation, thereby providing users with a more intelligent, coherent, and personalized interactive experience, has become a pressing technical problem to be solved in the current smart TV field.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0007] This disclosure provides a method, apparatus, smart TV, and computer-readable storage medium for generating favorites on a smart TV, offering users a smarter, more coherent, and more personalized interactive experience.

[0008] In some embodiments, the method for generating a favorites list for a smart TV includes: generating a favorites entry containing a unique identifier of the content in response to a user's trigger operation to favorite target media content; associating multi-dimensional context information with the favorites entry to construct a smart favorites unit; the multi-dimensional context information includes: timestamp information generated based on the time of the trigger operation, real-time scene mode information obtained through analysis of environmental sensors, and emotional state information obtained based on real-time analysis of the user's facial image; generating a composite operation protocol for one-click triggering of a preset function sequence in response to a combined function setting instruction for the target media content, and binding the composite operation protocol to the smart favorites unit; synchronizing the smart favorites unit to multiple terminal devices associated with the current user account so that the smart favorites unit remains callable on multiple terminal devices.

[0009] In some embodiments, the favorites generation apparatus for a smart TV includes a processor and a memory storing program instructions, the processor being configured to execute the aforementioned favorites generation method for a smart TV when running the program instructions.

[0010] In some embodiments, the smart TV includes: a smart TV body; and the aforementioned favorites generation device for the smart TV, which is installed on the smart TV body.

[0011] In some embodiments, the computer-readable storage medium stores program instructions that, when executed, cause a computer to perform the aforementioned method for generating favorites for a smart TV.

[0012] The method, apparatus, smart TV, and computer-readable storage medium for generating favorites on smart TVs provided in this disclosure can achieve the following technical effects: This application constructs intelligent collection units by associating multi-dimensional dynamic contextual information such as time, scene, and emotion, and binds them to a composite protocol that can trigger complex operation sequences with a single click. This upgrades the collection from a static recorder to an intelligent object that can perceive and adapt to the user's real-time state. Simultaneously, by synchronizing the complete intelligent collection unit to multiple associated terminal devices, it achieves real-time consistency and seamless continuation of collected content and state across devices, fundamentally solving the problem of inability to achieve state synchronization and scene continuation. Thus, it provides users with an intelligently perceptive, dynamically adaptable, and cross-device coherent collection system, enhancing the personalization and convenience of user operations.

[0013] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0014] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrations and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are shown as similar elements. The drawings are not to be scaled. And wherein: Figure 1 This is a schematic diagram of a method for generating a favorites list for a smart TV, provided in an embodiment of this disclosure; Figure 2 This is a schematic diagram of a method for obtaining real-time scene mode information provided in an embodiment of this disclosure; Figure 3 This is a schematic diagram of a method for obtaining emotional state information provided in an embodiment of this disclosure; Figure 4 This is a schematic diagram of a method for generating a composite operation protocol provided in an embodiment of this disclosure; Figure 5 This is a schematic diagram of a method for synchronizing a smart collection unit provided in an embodiment of this disclosure; Figure 6 This disclosure provides a schematic diagram of a favorites generation device for smart TVs. Detailed Implementation

[0015] To provide a more detailed understanding of the features and technical content of the embodiments of this disclosure, the implementation of the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this disclosure. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0016] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this disclosure described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0017] Unless otherwise stated, the term "multiple" means two or more.

[0018] In this embodiment of the disclosure, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0019] The term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0020] The term "correspondence" can refer to an association or binding relationship. The correspondence between A and B means that there is an association or binding relationship between A and B.

[0021] Combination Figure 1 As shown, optionally, this disclosure provides a method for generating a favorites list for a smart TV, including: S11: In response to the user's action of adding the target media content to their favorites, the smart TV generates a favorite entry containing the unique identifier of that content.

[0022] S12, the smart TV associates multi-dimensional contextual information with the collection items to build a smart collection unit; the multi-dimensional contextual information includes: timestamp information generated based on the time of the trigger operation, real-time scene mode information obtained through environmental sensor analysis, and emotional state information obtained based on real-time analysis of the user's facial image.

[0023] S13, in response to the combined function setting command for the target media content, the smart TV generates a composite operation protocol for triggering a preset function sequence with one click, and binds the composite operation protocol to the smart collection unit.

[0024] S14, the smart TV synchronizes the smart collection unit to multiple terminal devices associated with the current user account, so that the smart collection unit remains accessible on multiple terminal devices.

[0025] In this embodiment, the smart TV responds in real time to user commands to favorite currently playing or selected target media content, issued via voice commands or physical buttons on the remote control. This target media content includes broadcast television programs, online streaming video, game applications, or system settings menu items. Specifically, when the system detects that the user issues a voice command to "favorite this" or long-presses the dedicated favorite button on the remote control, a favorite operation is triggered. Subsequently, the system generates a favorite entry corresponding to the target media content. The core of this entry is a globally unique content identifier, typically encapsulated as a structured data object, such as a string or JSON object containing the content source platform, content type, and a unique ID within the platform, thus unambiguously pointing to that specific content instance within the system.

[0026] In this embodiment, after generating a collection item, the smart TV immediately collects and associates three key multi-dimensional contextual information to construct an intelligent collection unit. First, the system automatically records the precise moment the collection trigger operation occurs, generating a timestamp, such as 21:30 on October 27, 2023. This information directly represents the specific time point when the user generated the collection action. Second, it collects environmental audio data through a built-in microphone and obtains a list of devices currently connected to the network. Voiceprint and feature analysis is performed on the audio data. If multiple voice characteristics are identified and the number of simultaneously connected associated devices exceeds a preset threshold, the current mode is determined to be a family sharing mode; otherwise, it is determined to be a personal exclusive mode, thus generating real-time scene mode information. Third, the system captures the user's facial image through a camera, extracts key feature frames, and inputs them into a pre-trained emotion recognition convolutional neural network model for analysis. The model outputs the confidence level of the user's current emotional state, including pleasure, fatigue, and irritability. The system determines the emotional state with the highest confidence level as the current emotional state information. Ultimately, the system combines the aforementioned timestamp information, real-time scene mode information, and emotional state information into a set of dynamic tags, binding and encapsulating them with the previously generated collection entries to form a context-aware intelligent collection unit. In practical applications, for example, when a user mentions adding a movie to their collection in the living room on a weekend evening, the system not only records the movie identifier but also associates and stores the weekend evening time, the shared family scene mode, and the pleasant emotional state obtained through camera analysis. This makes the collection unit an intelligent entity that integrates specific time, space, and the user's emotional memory. In this way, by injecting a dynamic context-aware dimension into static collection content, the collection evolves from a simple address book into an intelligent memory system capable of understanding the user's environment and state.

[0027] Furthermore, after a user completes personalized settings for the target media content, the smart TV responds to the user's combined function setting command and initiates the generation process of a composite operation protocol. The system first parses the command, separating the multiple discrete function operations set by the user and the specific parameters corresponding to each operation. Then, strictly following the logical execution order of these function operations, the system encodes each function operation and its parameters into a structured data object that can be parsed and executed by the system. The core of this data object is a list of operation sequences, where each item clearly defines a unique function identifier and a matching set of parameter fields, thus forming a complete and ordered set of machine-readable instructions. Finally, the system encapsulates this structured data object into an independent, storable, and transmittable composite operation protocol. In practical applications, for example, when a user watches a movie and manually sets three functions—skipping the 90-second intro, enabling eye protection mode to reduce blue light by 30%, and adjusting the sound to cinema surround sound—when the user chooses to save this combination of settings, the system encapsulates these three operations and their parameters into a structured protocol. This protocol is a package containing an ordered list of actions, for example, encapsulating the three function nodes using JSON format. In this way, by condensing a series of complex manual settings by the user into a single smart protocol that can be triggered with one click, cross-application, multi-step function combinations can be reproduced instantly, completely changing the cumbersome interaction mode of traditional TVs that requires layer-by-layer operation.

[0028] In this embodiment, after the smart TV completes the construction of the smart collection unit, it initiates a process to synchronize it to multiple terminal devices associated with the current user account, ensuring that the unit remains complete and immediately available on all associated devices. This synchronization process is not a simple data copy, but a state mirroring and continuation based on distributed ledger technology. The system securely and reliably distributes the complete smart collection unit data, including collection entries, multi-dimensional context information, and complex operation protocols, to other terminal devices under the account, such as another bedroom TV or the user's personal tablet, through a decentralized synchronization mechanism. During this process, each terminal device maintains data consistency through a consensus mechanism, ensuring that updates on any device are synchronized to the entire device network. In practical applications, for example, if a user collects a documentary on the living room smart TV and sets a personalized viewing protocol including high-definition picture quality and transparent subtitle backgrounds, after synchronization is complete, the user can log in to the same account on the bedroom TV and directly see the smart collection unit in the collection list; after clicking, the bedroom TV not only starts playing the same documentary but also automatically applies all previously saved picture quality and subtitle settings, achieving seamless cross-device experience continuation. In this way, by building a decentralized and reliable multi-device status synchronization network, the collection and the personalized scenarios it carries are freed from the physical constraints of a single device, realizing the free flow and seamless connection of the user's digital experience in space.

[0029] The bookmark generation method for smart TVs provided in this disclosure constructs intelligent bookmark units by associating multi-dimensional dynamic context information such as time, scene, and emotion, and binds them to a composite protocol that can trigger complex operation sequences with one click. This upgrades the bookmark from a static record of paths to an intelligent object that can sense and adapt to the user's real-time state. Simultaneously, by synchronizing the complete intelligent bookmark unit to multiple associated terminal devices, real-time consistency and seamless continuation of bookmarked content and state are achieved across devices, fundamentally solving the problem of inability to achieve state synchronization and scene continuation. Thus, this provides users with an intelligently perceptive, dynamically adaptable, and cross-device coherent bookmarking system, enhancing the personalization and convenience of user operations.

[0030] Optionally, the smart TV can synchronize the smart collection unit to multiple terminal devices, and also includes: If the target terminal devices are heterogeneous devices from different brands, the device control commands in the composite operation protocol will be converted into a command format supported by the target device through the IoT protocol adaptation layer before synchronization or execution.

[0031] Understandably, to address the reality of multiple brands of smart devices in a user's home environment, when the system detects that the target synchronization terminal is a different brand from the current smart TV and supports different IoT control protocols, it will not simply synchronize the original commands. The system will first identify the brand and model of the heterogeneous device, and then activate the built-in IoT protocol adaptation layer. This adaptation layer, as a general command conversion engine, will parse and convert the control commands based on the source device standard, encapsulated in the composite operating protocol of the smart collection unit, into equivalent command formats and data packets supported by the target device manufacturer's publicly available protocols in real time. After this protocol conversion is completed, the adapted command set will be synchronized to the target device along with other data from the smart collection unit, or sent and executed directly when invoked. With this solution, regardless of the complexity of the device ecosystem in a user's home, the personalized scenarios generated based on this solution can be accurately reproduced and continued. For example, a user creates a "Cinema Mode" on a mainstream brand TV in the living room, which includes dimming Samsung smart lights. When this smart storage unit is synchronized to another brand TV in the bedroom, the system automatically converts the lighting control commands to the IKEA smart lighting protocol supported by the bedroom. This ensures that the one-click light-off effect is consistently achieved across different spaces and devices. This design, by introducing a protocol adaptation layer, fundamentally breaks down the technical barriers between different brand devices, giving the smart storage system true cross-ecosystem compatibility and greatly expanding its application scope and user value.

[0032] Combination Figure 2 As shown, optionally, the real-time scene mode information obtained by the smart TV through analysis of environmental sensors includes: S21, Smart TV collects ambient audio data and a list of currently connected devices.

[0033] S221, if the environmental audio data indicates the presence of multiple voices and the number of associated devices is greater than a threshold based on the current list of connected devices, the smart TV determines the real-time scene mode information as the family sharing mode.

[0034] S222, if the environmental audio data indicates the presence of single-person voice characteristics and / or the number of associated devices is less than or equal to a threshold based on the current list of accessed devices, the smart TV determines the real-time scene mode information as a personal exclusive mode.

[0035] In this solution, the smart TV uses its built-in environmental sensors to intelligently analyze real-time scene modes. Specifically, the system first collects audio data from the current environment via a microphone, and simultaneously obtains a list of devices currently connected and active through the same home network and associated accounts. The system performs real-time acoustic analysis on the collected environmental audio data, focusing on identifying whether it contains multiple independent human voices with different timbres and tones, thereby determining whether there are voice characteristics of multiple people talking or active simultaneously. At the same time, the system counts the number of valid devices in the currently connected device list, such as smartphones, tablets, and smart speakers logged into the same account. The system performs a comprehensive logical judgment based on the information from the above two dimensions: if the environmental audio analysis indicates the presence of clear multi-person voice characteristics, and the number of currently connected associated devices exceeds a preset threshold, such as two or more, the system determines that the current environment is shared by multiple people, thus identifying the real-time scene mode as a home-sharing mode; conversely, if the environmental audio analysis indicates the presence of only single-person voice characteristics, or the number of currently connected associated devices does not exceed the threshold, or both conditions are met, the system determines that the current environment is a personal use environment, thus identifying the real-time scene mode as a personal exclusive mode. In practical applications, for example, when family members gather in the living room at night, the TV microphone will capture the sounds of multiple people talking and laughing, and simultaneously detect that multiple family members' mobile phones are connected to the home Wi-Fi. The system will then automatically mark this scenario as a family sharing mode. Conversely, when only one user is watching a program late at night and all other devices are asleep, the system will determine it as a personal viewing mode. In this way, by integrating environmental sound characteristics and network device status for collaborative analysis, the system can dynamically and accurately perceive the user's physical and social context, providing crucial contextual basis for subsequent intelligent content filtering and personalized recommendations.

[0036] In one optimized solution, the smart TV analyzes real-time scene mode information using environmental sensors, and also includes: Smart TVs collect ambient light intensity data and thermal maps of human movement within the space.

[0037] Based on environmental audio data, ambient light intensity data, and human movement heatmaps, smart TVs use a multimodal fusion model to determine whether the current scenario is a family sharing mode, a personal focus mode, a party entertainment mode, or a nighttime rest mode.

[0038] In this solution, an additional environmental perception dimension is introduced to improve the accuracy and richness of scene recognition. In addition to collecting ambient audio and a list of connected devices, real-time ambient light intensity data is simultaneously acquired via a light sensor. Furthermore, a human movement heatmap representing the distribution and activity intensity of people within the space is generated using a built-in visual sensor or infrared sensor array. Subsequently, the system inputs these four types of heterogeneous sensor data—ambient audio data, ambient light intensity data, human movement heatmap, and the current list of connected devices—into a pre-trained multimodal fusion analysis model for collaborative analysis and decision-making. This model can deeply understand the correlations between different data modalities. For example, combining dim lighting, a low-activity heatmap, quiet audio, and a few online devices can accurately identify a nighttime rest mode; while combining bright lighting, a high-activity and dispersed heatmap, noisy multi-person audio, and multiple online devices may indicate a party / entertainment mode. In practical applications, when the TV detects soft ambient light, a heatmap showing a single person stationary in front of the screen, quiet audio, and only the TV itself online, the system can accurately identify it as a personal focus mode rather than a simple personal enjoyment mode, even if other devices are on the network. This solution, by integrating visual, acoustic, light-sensing, and network information, achieves a leap from simple binary classification to refined categorization of user scenarios, significantly improving the depth and reliability of contextual awareness. This lays a more solid foundation for providing highly personalized content and services that perfectly match the current atmosphere.

[0039] In one example, a pre-trained multimodal fusion analysis model receives and processes standardized input data from different sensors: converting ambient audio data into feature vectors containing volume, spectrum, and probabilities of specific sound categories such as speech, music, and laughter; normalizing ambient light intensity data into a numerical value representing brightness levels; compressing the pixel matrix of human movement heatmaps into a low-dimensional feature vector representing the number of people, their location distribution, and their activity level; and processing the list of access devices into encoded sequences of device quantity and type. Through its internal attention mechanism and feature fusion layer, the model learns complex correlation patterns between these heterogeneous features. For example, it might learn to highly correlate and classify a series of features such as high brightness, multiple voices, dispersed hotspot distribution, and multiple online entertainment devices as a party entertainment mode. This model is trained on a large amount of labeled scene data and can ultimately output a classification probability of the current scene belonging to one of four categories: family sharing mode, individual focus mode, party entertainment mode, or nighttime rest mode, based on the real-time multimodal sensor data stream. This achieves a highly robust and refined scene understanding that far surpasses the logical judgment of a single sensor.

[0040] Optionally, the core design of the multimodal fusion analysis model includes a multi-head attention mechanism layer for cross-modal interaction and a multi-layer feature fusion network for generating the final decision. Specifically, after preprocessing, the audio features, light intensity scalar, thermal feature vectors, and device encoding vectors from each sensor are mapped to the same high-dimensional semantic space, serving as independent feature sequences input to the model. The multi-head attention mechanism layer allows the feature vectors of each modality to interact and compute with the feature vectors of all other modalities. For example, it calculates the attention weights between audio and thermal features, thereby dynamically capturing cross-modal associations such as whether a loud sound overlaps with the thermal regions of multiple people. Subsequently, these attention-weighted and interacted enhanced features are fed into the feature fusion network, which typically consists of multiple fully connected layers and nonlinear activation functions. This network is responsible for deep abstraction and dimensionality reduction of the fused high-dimensional features, and finally outputs the probability distributions corresponding to the four scene modes through a Softmax classification layer. In a trained model instance, when the input features include moderate lighting, detection of high-frequency child voice features, a heatmap showing two adjacent hotspots, and a device list containing a game console, the attention mechanism might assign higher weights between the child's voice and the two hotspots. The fusion network, however, learns from historical data the strong association between this combination and family sharing patterns, thus outputting a high-confidence value for the pattern. This design allows the model to simulate the human cognitive process of comprehensively judging multiple cues, rather than relying on a single signal, thereby significantly improving the accuracy and robustness of scene classification in complex real-world home environments.

[0041] Combination Figure 3 As shown, the emotional state information obtained by the smart TV based on real-time analysis of the user's facial images includes: The S31 smart TV captures a video stream containing the user's face via a camera.

[0042] S32, the smart TV extracts facial feature frames from the video stream and inputs them into a pre-trained emotion recognition convolutional neural network model.

[0043] S33, the smart TV acquires the output of the convolutional neural network model, which includes multiple emotional states such as pleasure, calmness, fatigue, and irritability, as well as the confidence level corresponding to each emotional state.

[0044] S34, the smart TV identifies the emotional state with the highest confidence level as the emotional state information.

[0045] In this embodiment, the smart TV obtains emotional state information based on real-time analysis of the user's facial images. Specifically, the system first continuously captures a real-time video stream facing the viewing area and containing the user's face through an integrated or external camera. The system extracts clear, frontal facial images from the video stream at certain frame intervals as key feature frames. Subsequently, these preprocessed facial feature frames are input into an emotion recognition convolutional neural network model pre-trained on massive amounts of facial emotion data for inference analysis. The convolutional neural network model performs deep processing on the input facial features and finally outputs a probability distribution describing the user's current emotional state. This distribution covers multiple preset emotion categories such as pleasure, fatigue, and irritability, and provides a confidence score for each category. After receiving this probability distribution, the system does not simply select the emotion with the highest confidence, but applies decision logic: if the score corresponding to the emotion category with the highest confidence exceeds a preset emotion judgment threshold, the system determines that this emotional state is the final emotional state information; if the score with the highest confidence does not reach the threshold, the system may determine the current emotional state as unknown or maintain the previous valid state for conservative considerations. As an example, the emotion judgment threshold is 0.7. In practical applications, for instance, when a user is watching a program late at night, the camera captures their slightly sleepy face. After analysis by a convolutional neural network model, the system outputs a fatigue confidence score of 0.85 and a calm confidence score of 0.12. Since the fatigue confidence score is significantly higher than the preset threshold, the system determines the current emotional state as fatigue. In this way, by combining non-contact visual analysis with the precise classification and threshold decision-making mechanism of deep learning, the system can perceive the user's emotional state in real time and objectively, laying the perceptual foundation for providing intelligent content recommendations and interactions that match the user's psychological needs.

[0046] Optionally, the smart TV, based on the emotional state information obtained from real-time analysis of the user's facial images, also includes: Smart TVs collect users' voiceprint characteristics and tone of voice information through microphones.

[0047] Smart TVs integrate facial image analysis results with voiceprint features and tone information, and through decision-level fusion, obtain the final emotional state information.

[0048] In this embodiment, a multimodal perception and fusion mechanism is introduced to improve the comprehensiveness and fault tolerance of emotional state judgment. While analyzing facial images captured by a camera, the system simultaneously collects user speech segments using a built-in microphone and extracts voiceprint features characterizing individual voice traits, such as formants, and intonation information reflecting immediate emotional changes, such as pitch, speech rate, and energy variation curves. Subsequently, the system inputs the emotional probability distribution obtained from facial image analysis and the independent emotional tendency judgment obtained from acoustic analysis into a decision-level fusion module. This module does not simply average the results but follows a preset fusion strategy. For example, when insufficient lighting leads to low confidence in facial features, the weight of acoustic features is increased to comprehensively evaluate the two types of evidence, ultimately generating a unified and reliable emotional state judgment. For instance, when a user says "continue playing" in a tired tone, even if their face is not identified as tired with high confidence by the model due to dim lighting, the system can still accurately determine the final emotional state as tired by fusing the low-pitched tone features. This solution effectively overcomes the limitations of single face recognition in complex scenarios such as poor lighting and user profiles by integrating visual and auditory cues, significantly enhancing the robustness and environmental adaptability of the emotion perception system, thereby providing users with a more accurate and stable foundation for emotional interaction.

[0049] Optionally, S34, the smart TV identifies the emotional state with the highest confidence level as the emotional state information, including: The smart TV determines whether the confidence level corresponding to the emotional state with the highest confidence level is greater than a preset emotional judgment threshold. If so, the emotional state is identified as emotional state information; otherwise, it is determined as an unknown emotional state.

[0050] In this scheme, the system pre-sets an emotion judgment threshold, which represents the minimum confidence level at which the model's output can be trusted. After obtaining the emotion probability distribution output by the convolutional neural network model, the system first identifies the emotion category with the highest confidence score and then compares its score with the pre-set threshold. If the highest confidence score exceeds the threshold, the system adopts this emotion state as the final emotion state information; conversely, if the highest confidence score does not reach the threshold, for example, if the confidence scores of multiple emotion categories are low and not significantly different, the system considers the evidence provided by the current image analysis insufficient and marks the current emotion state as unknown. In practical applications, this design effectively prevents the system from making arbitrary and potentially erroneous emotion judgments when the user's face is partially obscured, lighting conditions are extremely poor, or the user's expression is neutral, leading to generally low confidence scores in the model output. For example, when a user is turned to the side to talk to a family member, the facial image captured by the camera is incomplete. The model outputs results such as "pleasure" (0.45) and "calm" (0.41), which are not clearly dominant and have a confidence level below the 0.5 threshold. In this case, the system will determine the emotional state as unknown, thus avoiding inappropriate recommendations of cheerful content that might be based on incorrect pleasure judgments. This approach significantly improves the accuracy and reliability of the emotion perception system, ensuring that only high-quality, high-confidence emotion recognition results drive subsequent personalized services, effectively enhancing the user experience's fit and comfort.

[0051] Combination Figure 4 As shown, optionally, in step S13, the smart TV, in response to a combined function setting instruction for the target media content, generates a composite operation protocol for one-click triggering of a preset function sequence, including: S41 is a smart TV parsing and combination function setting command that extracts multiple discrete function operations and their corresponding parameters.

[0052] S42, the smart TV encodes each function operation and its corresponding parameters into a structured data object according to the execution order of discrete function operations. The data object includes an operation sequence list, and each item in the list is defined with a function identifier and its parameter field.

[0053] S43, smart TVs encapsulate structured data objects into composite operating protocols.

[0054] In this embodiment, in response to a user's combined function setting command for target media content, the smart TV first parses the user's setting command, precisely extracting multiple discrete function operations that the user intends to execute and the specific parameter values ​​corresponding to each operation. These function operations can broadly cover various types, such as controlling media content playback (e.g., skipping specific segments), adjusting image parameters of the display device (e.g., color temperature, brightness), adjusting output parameters of the audio device (e.g., sound effect modes, equalizers), and even controlling smart home devices associated with the smart TV (e.g., smart lights, curtains). Subsequently, the system strictly follows the logic and time sequence of the user's settings, encoding each function operation and its parameters into a data object with a strict internal structure. The core of this data object is an ordered list of operation sequences, where each item clearly defines a function identifier to uniquely identify the function, as well as a parameter field carrying its specific setting value. Finally, the system encapsulates this structured data object into a complete, independently storable, and transmittable composite operation protocol. In practical applications, for example, when a user creates an immersive viewing mode for a movie, they might sequentially set options such as skipping the 90-second intro, enabling cinema-grade color temperature and contrast, activating Dolby Atmos, and adjusting smart lighting to the viewing scene. The system then encodes these four different types and parameters into a structured protocol packet according to their setting order. In this way, by transforming the complex, manual sequence of settings across applications and devices into a persistent and precisely reproducible structured instruction set, a fundamental shift from cumbersome multi-step operations to efficient one-click triggering is achieved.

[0055] Optionally, the function operations in the preset function sequence include at least one of the following types: Media content playback control operations.

[0056] Display device image parameter adjustment operation.

[0057] Adjusting audio device output parameters.

[0058] Controlling smart home devices associated with a smart TV.

[0059] In this solution, a pre-defined sequence of functions can be encapsulated within a composite operation protocol. This sequence encompasses a wide range of highly integrated functional operations. Specifically, these operations extend beyond controlling the television itself to the entire home entertainment and living environment it interacts with. These operations primarily include: controlling media content playback behavior itself, such as skipping, speed adjustment, and looping; adjusting image processing parameters of smart TVs or associated display devices, such as switching picture quality modes, adjusting color temperature and dynamic contrast; adjusting audio system or audio output parameters, such as switching sound field modes, adjusting equalizers and volume equalizers; and controlling the status of smart home devices connected to the television via IoT protocols, such as smart lighting, air conditioning, and curtains. In actual protocol generation, a single protocol can simultaneously include multiple types of operations. For example, a home theater protocol might encapsulate a series of cross-domain operations, such as starting movie A, switching to movie picture quality mode, enabling Dolby Atmos, and dimming the living room lights to 20%. By expanding the scope of control from single media playback to display, audio, and even the entire associated physical environment, a single collection unit can trigger a complete and coordinated immersive scene with a single click, enriching the application depth and user experience of smart collections.

[0060] In one example, the process of creating combined function settings commands can be achieved through a visual function chain editor. When a user triggers an advanced collection or scene creation command for target media content, the system presents an editing interface. This interface displays the currently playing content on one side and provides an expandable function pool panel on the other, listing discrete function operations that can be added, such as skipping the intro, setting the picture quality mode, adjusting the volume equalizer, and controlling smart lights. The user drags and drops the desired functions from the function pool into the editing area, forming an ordered chain. For each added function, the system pops up a parameter setting panel; for example, setting specific start and end timestamps for the skip intro function and selecting a cinema preset for the picture quality mode. After the user completes all function additions and parameter configurations, and confirms and saves, the system then triggers the protocol generation process S41-S43 as described above, solidifying this series of interactive operations into a storable and retrievable composite operation protocol. This interactive design allows non-technical users to easily define complex cross-application, cross-device automation scenarios. Combination Figure 5 As shown, optionally, in S14, the smart TV synchronizes the smart collection unit to multiple terminal devices associated with the current user account, including: S51: When a shared data ledger has been established between the smart TV and multiple terminal devices, the smart TV generates corresponding data transaction records.

[0061] S52, the smart TV distributes data transaction records to multiple terminal devices associated with the current user account, so that the multiple terminal devices can perform consensus verification. Consensus verification includes verifying the digital signature of the data transaction records and verifying that the order of the data transaction records in the shared data ledger conforms to preset rules.

[0062] S53, after receiving the consensus verification result from the associated terminal device, the smart TV appends the data transaction record to the smart TV's shared data ledger and instructs the associated multiple terminal devices to append it synchronously, so as to realize the state synchronization of the smart collection unit.

[0063] In this embodiment, assuming a shared data ledger has been pre-built and maintained by the smart TV and multiple associated terminal devices, when the smart TV needs to synchronize a new or updated smart collection unit, it first generates a corresponding data transaction record based on all data of that unit, including content identifiers, multi-dimensional context information, and composite operation protocols. This record includes the operation type, data content, and a digital signature generated by the smart TV's private key. Subsequently, the smart TV broadcasts and distributes this data transaction record to all other terminal devices associated with the current user account. Each terminal device receiving the record independently performs consensus verification, mainly including two checks: first, verifying the digital signature attached to the record to ensure that the record was indeed issued by the legitimate source device, i.e., the smart TV, and has not been tampered with; second, verifying whether the order of this record in the historical record of the shared data ledger conforms to preset rules, such as timestamp incrementing, to prevent data conflicts and state rollback. Once the smart TV collects consensus verification results from a sufficient number of associated terminal devices, it considers the transaction to have achieved network consensus. It then formally appends this data transaction record to its own maintained shared data ledger and simultaneously instructs all associated terminal devices to append the same record to their respective local ledger copies. This ensures consistency in the ledger state across all devices, ultimately achieving synchronization and state consistency of the smart collection unit. In practical applications, when a user creates a smart collection unit with specific viewing settings on the living room smart TV, the system initiates the above synchronization process. After receiving the transaction record and completing signature and sequence verification, the bedroom TV and tablet update their respective ledgers, making the collection unit immediately appear in their favorites and fully accessible. Thus, by introducing a decentralized synchronization mechanism based on a shared ledger and consensus verification, not only is real-time consistency of collection data achieved across multiple devices, but the reliability and security of data transmission are also fundamentally guaranteed. This completely breaks down the data silos and single-point dependencies of traditional centralized synchronization models, building a truly coherent and reliable cross-device smart experience network for users.

[0064] In a specific and efficient implementation architecture, state synchronization among multiple terminal devices can be achieved through a centralized synchronization service based on user accounts. Each terminal device's client, after generating or modifying a smart collection unit locally, encapsulates its complete data and an incremental operation command (such as create, update, or delete) into a synchronization message and sends it to the central cloud server bound to the user account. The cloud server, acting as the authoritative node, records the synchronization message in a globally ordered operation log in the order of receipt and immediately pushes it to all other online terminal devices under that account. Upon receiving the synchronization message, each terminal device applies the operation locally, updating its own collection state. To handle network interruptions or conflicts, the cloud server maintains a global version number for each smart collection unit and, upon detecting a conflict, adopts a strategy of either last-write-wins or initiating conflict resolution with the user. This solution ensures eventual data consistency while avoiding the complex consensus overhead of distributed ledgers, making it more suitable for home IoT environments.

[0065] Optionally, the shared data ledger records the calling status mapping relationship of the smart collection unit across multiple terminal devices; the method further includes: In response to a smart TV invoking a smart collection unit on the first terminal device, the smart TV synchronously updates the status identifier of the smart collection unit in the corresponding user interface on the second terminal device according to the invoking status mapping relationship.

[0066] In this embodiment, the shared data ledger, in addition to recording the static data of smart collection units, also maintains a dynamic call status mapping relationship. This mapping relationship tracks and associates the current call status of each smart collection unit on various terminal devices in real time, such as pending playback, playing, or paused. When a user calls a smart collection unit on the first terminal device, this call event itself will also generate a special data transaction record that updates the call status mapping relationship, and synchronize it to the ledger network via the aforementioned consensus mechanism. After the ledger is updated, other associated terminal devices will actively refresh their local user interface according to this new mapping relationship, updating the status identifier of the corresponding smart collection unit. In practical applications, when a user clicks on and starts playing a collected weekend animation time slot on the living room TV, the living room TV immediately generates a status update transaction; after the transaction is verified by consensus and written into the shared ledger, a dynamic identifier or icon indicating that it is playing in the living room will automatically and in real time be displayed next to the corresponding collection item on the tablet computer in the children's room. In this way, by treating users' real-time interactive behaviors as synchronized status information, the collection system not only achieves seamless migration of static content, but also makes dynamic usage scenarios transparent and visible. This allows multiple devices in the home to truly perceive each other's usage status, providing users with a sense of global control and scenario continuity. This elevates the cross-device collaborative experience from simple data synchronization to a new level of intelligent dynamic behavior linkage.

[0067] In one optimized approach, the method further includes: When a network connection interruption is detected between the smart TV and some terminal devices, data transaction records are recorded locally. Once the network is restored, the cached data transaction records are automatically appended to the local shared data ledger, and an attempt is made to synchronize with the terminal devices that have regained connection.

[0068] In this solution, when a smart TV is performing a synchronization process, if it detects a temporary interruption in the network connection with one or more associated terminal devices, it will not cause the entire synchronization operation to fail or enter a blocked state. Instead, the system maintains an offline queue for synchronization locally, securely caching data transaction records that failed to be distributed due to the network interruption. When the network connection is restored, the system automatically detects this state change and immediately retrieves the previously cached transaction records from the offline queue, appends them to the local shared data ledger, and actively attempts to re-initiate the synchronization process with the now-reconnected terminal devices. In practical applications, when a user modifies a favorite unit on the master bedroom TV and triggers synchronization, and the children's room TV happens to go offline due to network fluctuations, the master bedroom TV will cache the update transaction record locally. After the children's room TV's network is restored, the system will automatically complete the data synchronization, ensuring that the latest modifications take effect on the children's room TV. This solution introduces offline caching and breakpoint resume capabilities, ensuring that even in the face of unstable or unreliable home network environments, the smart collection system's state synchronization can still maintain eventual consistency and operational continuity, greatly improving the system's fault tolerance and user experience reliability in actual deployment.

[0069] Optionally, after synchronizing the smart collection unit to multiple terminal devices associated with the current user account, the method further includes: The smart TV obtains the current real-time context information.

[0070] The smart TV matches and calculates real-time contextual information with the multi-dimensional contextual information associated with the smart collection unit.

[0071] Based on the matching calculation results, the smart TV sorts the multiple smart collection units that are currently available.

[0072] In this solution, the smart TV first acquires real-time context information through its sensor array. This information includes real-time time information, current scene mode information obtained through environmental sensor analysis, and emotional state information obtained based on the current user's facial image analysis. The system then performs a matching calculation on this set of real-time context information with the multi-dimensional context information associated with and stored when constructing each smart collection unit. This matching calculation is typically performed using a matching degree algorithm, which compares the similarity or correlation strength between the real-time information and the historical information of each smart collection unit across various dimensions, and calculates an overall matching degree score. Based on this matching calculation result, the system reorders all available smart collection units on the current device, placing units with higher matching degrees in higher priority positions. In practical applications, for example, when the system detects a family-shared scene on a weekend evening and the user's emotional state is pleasant, it automatically calculates the matching degree of all collection units with tags such as weekend, family, and pleasant, and promotes a smart collection unit of children's animations historically collected in a similar weekend family-happy scene to the top of the list, while placing a work lecture unit marked as personal enjoyment or collected when tired to the bottom. In this way, by intelligently matching and dynamically rearranging the user's real-time status with the historical context of the saved content, the system can proactively and accurately recommend saved content, allowing users to quickly find the saved content that best matches their current mood and scenario without having to search manually, greatly improving the proactivity of human-computer interaction and the efficiency of personalized services.

[0073] In one example, the system pre-sets weights and scoring rules for three dimensions: time, scenario, and emotion. For instance, the weight for the time dimension is set to 0.3, the weight for the scenario dimension to 0.5, and the weight for the emotion dimension to 0.2. During scoring, if the current time, such as Friday 8:00 PM, matches the time tag recorded in the collection unit, such as a weekend evening, the time dimension receives a perfect score of 100; if it partially matches, such as a weekday evening, it receives 60 points; and if it doesn't match at all, it receives 0 points. In the scenario dimension, if the current scenario, such as family sharing, completely matches the historical tag, it receives a perfect score of 100; a compatible scenario, such as individual sharing, receives 50 points; and a conflicting scenario, such as family sharing and individual sharing, receives 0 points. In the emotion dimension, a completely consistent state receives 100 points; a positive state, such as pleasant and calm, receives 70 points; and a negative state, such as pleasant and tired, receives 0 points. The final matching score is a weighted sum. In practical applications, when a family gathering takes place on Friday night, a variety show compilation tagged as "Weekend," "Family," and "Enjoyable" might score as follows: Time 100 * 0.3 + Scene 100 * 0.5 + Emotion 100 * 0.2 = 100 points. Conversely, a documentary tagged as "Tuesday Night," "Personal," and "Tired" might score as follows: Time 60 * 0.3 + Scene 0 * 0.5 + Emotion 0 * 0.2 = 18 points. The system then ranks the variety show compilations accordingly. This quantitative algorithm, through explicit rules and numerical calculations, transforms scenario matching into an objective and repeatable ranking criterion, ensuring the transparency and consistency of the dynamic recommendation logic.

[0074] In one optimized approach, the smart TV sorts the multiple currently available smart collection units based on the matching calculation results, including: If the current real-time context information indicates a family sharing scenario, the priority of smart collection units associated with personal private content will be reduced in the ranking; if the current real-time context information indicates that the user is in a tired or irritable emotional state, the priority of smart collection units associated with soothing and relaxing media content will be increased in the ranking.

[0075] In this solution, when the system identifies a family-shared scenario based on real-time contextual information, regardless of other matching dimensions, the ranking algorithm proactively applies a priority penalty coefficient to smart collection units marked as personal private content, significantly lowering their position in the list. Simultaneously, if the system analyzes that the user is currently in a tired or irritable emotional state, the algorithm assigns an additional priority boost coefficient to all smart collection units associated with soothing and relaxing types, regardless of the original content category, causing their ranking to jump. In practical applications, for example, when the system detects a family gathering on a weekend night and the female homeowner appears tired, the algorithm will, on the one hand, strongly suppress highly private units such as the male homeowner's private work files, preventing them from appearing in the top recommendations; on the other hand, it will significantly boost the ranking of a previously collected set of forest healing videos tagged as tranquil and natural, prioritizing its presentation. This approach, by introducing strong scenario rules and emotion-oriented weighting interventions, enables dynamic ranking to not only be based on similarity but also to deeply align with the actual needs of privacy protection and emotional care. It achieves the evolution of intelligent recommendation from objective relevance to subjective suitability, significantly enhancing the system's service thoughtfulness and scenario security in different family roles and user states.

[0076] Optionally, the method may further include: when the real-time scene mode information is determined to be in family sharing mode, the system will access a predefined content security policy library, which defines the visibility rules of different content tags in specific scenes. Based on these rules, the system will iterate through the currently available smart collection units: if the content metadata associated with a smart collection unit contains restricted tags such as personal privacy or adult content, the system will directly remove the unit from the main display list of the current user interface, or display it as an unclickable grayscale state, and optionally indicate that the content is unavailable in the current scene. For example, in an evening family gathering scene where a child's voice is detected, work videos that the user has individually collected and marked as private files will be automatically hidden to ensure they are not accidentally displayed on the family's public screen. In this way, by combining scene awareness with mandatory filtering rules for content tags, proactive and rigid privacy protection is achieved at the system level, mitigating the risk of accidental touches that may arise from relying solely on ranking and demotion.

[0077] Optionally, the method further includes: The smart TV responds to the user's command to call the smart collection unit and obtains the composite operating protocol bound to the smart collection unit.

[0078] The smart TV analyzes the composite operation protocol to obtain the preset function sequence.

[0079] The smart TV controls the smart TV to execute a preset function sequence to trigger the playback of target media content and simultaneously execute multiple function operations encapsulated in the preset function sequence.

[0080] In this solution, when a user selects and issues a command to call a smart collection unit on the smart TV's collection interface, the system first retrieves the pre-bound, complete composite operation protocol from that unit. The system then parses the protocol, decoding the pre-defined sequence of functions encapsulated in sequence. This sequence specifies all the functions to be executed sequentially and their precise parameters. Next, the smart TV, as the control core, begins executing this sequence line by line: it first triggers the playback of the target media content based on the content identifier indicated in the protocol, and simultaneously, or in a predetermined order, automatically and continuously executes all other encapsulated functions within the sequence. These operations can span multiple control domains; for example, at the moment playback begins, the system may simultaneously adjust the local picture and sound parameters and send control commands to associated smart home devices. In practical applications, when a user calls a smart collection unit called "Starry Sky Viewing Mode," the system, after parsing its protocol, immediately begins playing the specified science fiction movie and almost simultaneously adjusts the TV screen's contrast and color temperature to the preset starry sky viewing parameters, switches the sound system to immersive surround mode, and dims and changes the living room's smart lights to blue ambient light via IoT commands. In this way, by parsing the call instruction into atomic execution of the composite operation protocol, a precise mapping is achieved from a single user interaction to a cross-domain, multi-step collaborative response of the system.

[0081] Optionally, before execution, the system verifies whether the current user or the current scenario has the permission to perform specific sensitive operations within a preset function sequence. Here, specific sensitive operations refer to payment or access to private content. If permission is not granted, execution is paused or replaced with a secure operation.

[0082] In this solution, the system first identifies whether the sequence contains instructions marked as specific sensitive operations, such as payment transactions or access to a user's private storage. Once such a sensitive operation is identified, the system immediately suspends the execution of all subsequent instructions and initiates a permission verification process. This process comprehensively verifies the current user's identity and the current real-time scenario. Only when the permission verification passes, confirming that the current context has the legal authority to execute the sensitive operation, will the system continue executing the complete sequence. If the verification fails, the system will take security measures, such as completely suspending execution and displaying a permission deficiency warning, or automatically replacing the sensitive operation with a safe default operation. In practical applications, when a child accidentally taps an animated series collection unit containing a purchase option for the next episode in a family-shared scenario, the system, after recognizing the payment instruction, the child's identity, and the family scenario, will prevent payment and only play the currently free episode, while simultaneously displaying a message on the interface indicating that the purchase function is disabled. This secure call mechanism ensures a balance between convenience and security by embedding an intelligent permission gateway into a convenient one-click trigger process. It effectively prevents risks of accidental operation, unauthorized access, or privacy leaks, and greatly improves the system's reliability and user trust in complex home multi-user environments.

[0083] In one optimized approach, the method further includes: The smart device records the frequency of invocation of the smart collection unit across multiple terminal devices and the context information at the time of invocation. Based on the invocation frequency and context information, the smart device dynamically weights or adaptively corrects the multi-dimensional context information associated with the smart collection unit. Adaptive correction includes: if a smart collection unit is frequently invoked under a specific emotional state, increasing the weight of that emotional state information in the multi-dimensional context information; and / or, if a smart collection unit is frequently invoked within a specific time period, adding or strengthening the time dimension label corresponding to that time period.

[0084] In this solution, the system continuously records historical access data for each smart collection unit across all associated terminal devices, including the frequency of its access and the real-time context information recorded by the system at each access. The system periodically analyzes this accumulated data and, based on the analysis results, dynamically and data-drivenly optimizes the static multidimensional context information tags associated with the corresponding smart collection units—a process known as adaptive correction. For example, if the system analysis finds that a classical music collection unit is consistently accessed when the user's emotional state is identified as fatigued, even if its original tag does not include fatigue, the system will automatically increase the weight of the fatigued emotional state in its multidimensional context information or directly add that emotional tag. Similarly, if a children's animation unit is frequently accessed on weekday evenings, the system may add or strengthen the weekday evening time dimension tag for it. In practical applications, a collection of game videos initially saved by a user casually on a weekend afternoon might, after a period of monitoring, be found to be clicked more frequently during the user's irritable weekday lunch break. The system then adaptively adjusts its associated information, increasing the weight of irritability and weekday lunchtimes, making its future ranking and recommendations more accurate for the same time period and mood. This allows the intelligent collection system to transcend initial, static tags based on a single snapshot, continuously refining and enriching its understanding of user intent and context through real-world usage feedback, significantly improving personalization and system intelligence over long-term use.

[0085] Combination Figure 6 As shown, this disclosure provides a favorites generation device 300 for smart TVs, including a processor 301 and a memory 302. Optionally, the device 300 may further include a communication interface 303 and a bus 304. The processor 301, communication interface 303, and memory 302 can communicate with each other via the bus 304. The communication interface 303 can be used for information transmission. The processor 301 can call logical instructions in the memory 302 to execute the favorites generation method for smart TVs described in the above embodiment.

[0086] Furthermore, the logic instructions in the aforementioned memory 302 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0087] The memory 302, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor 301 executes functional applications and data processing by running the program instructions / modules stored in the memory 302, thereby implementing the bookmark generation method for smart TVs in the above embodiments.

[0088] The memory 302 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 302 may include high-speed random access memory and may also include non-volatile memory.

[0089] This disclosure provides a smart TV, including a smart TV body and the aforementioned bookmark generation device 300 for the smart TV. The bookmark generation device 300 is mounted on the smart TV body. The mounting relationship described herein is not limited to placement inside the smart TV body, but also includes mounting connections with other components of the smart TV, including but not limited to physical connections, electrical connections, or signal transmission connections. Those skilled in the art will understand that the bookmark generation device 300 for the smart TV can be adapted to any feasible smart TV body, thereby realizing other feasible embodiments.

[0090] This disclosure provides a computer-readable storage medium storing computer-executable instructions configured to perform the above-described method for generating a favorites list for a smart TV.

[0091] The technical solutions of this disclosure can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in this disclosure. The aforementioned storage medium can be a non-transitory storage medium, such as a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, etc., and other media capable of storing program code.

[0092] The foregoing description and accompanying drawings fully illustrate embodiments of this disclosure to enable those skilled in the art to practice them. Other embodiments may include structural, logical, electrical, procedural, and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the terminology used in this application is for describing embodiments only and is not intended to limit the claims. As used in the description of embodiments and claims, the singular forms “a,” “an,” and “the” are intended to equally include the plural forms unless the context clearly indicates otherwise. Similarly, the term “and / or” as used in this application means including one or more of the associated listed items and all possible combinations thereof. Additionally, when used in this application, the term "comprise" and its variations "comprises" and / or "comprising" refer to the presence of stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. Without further limitations, an element defined by the phrase "comprises a..." does not exclude the presence of other identical elements in the process, method, or apparatus that includes said element. In this document, each embodiment may focus on the differences from other embodiments, and similar or identical parts between embodiments can be referred to mutually. For methods, products, etc., disclosed in the embodiments, if they correspond to the method section disclosed in the embodiments, the relevant parts can be referred to the description of the method section.

[0093] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. 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 implementation should not be considered beyond the scope of the embodiments of this disclosure. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0094] The methods and products disclosed in the embodiments herein (including but not limited to devices and equipment) can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units may be merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple 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 through some interfaces, and the indirect coupling or communication connection of devices or units may be electrical, mechanical, or other forms. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to implement this embodiment according to actual needs. In addition, the functional units in the embodiments of this disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0095] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

Claims

1. A method for generating a favorites list for a smart TV, characterized in that, include: In response to a user's action of adding a target media content to their favorites, a favorite entry containing the unique identifier of that content is generated. Multidimensional contextual information is associated with the collection entries to construct intelligent collection units; The multidimensional context information includes: timestamp information generated based on the time of occurrence of the triggering operation, real-time scene mode information obtained through environmental sensor analysis, and emotional state information obtained based on real-time analysis of the user's facial image; In response to the combined function setting instruction for the target media content, a composite operation protocol is generated for triggering a preset function sequence with one click, and the composite operation protocol is bound to the smart collection unit; The smart collection unit is synchronized to multiple terminal devices associated with the current user account so that the smart collection unit remains callable on the multiple terminal devices.

2. The method according to claim 1, characterized in that, Real-time scene pattern information obtained through environmental sensor analysis includes: Collect environmental audio data and a list of currently connected devices; If the environmental audio data indicates the presence of multiple voices and the number of associated devices is greater than a threshold based on the current list of connected devices, the real-time scene mode information is determined to be a home sharing mode. If the environmental audio data indicates the presence of single-person voice characteristics and / or the number of associated devices is less than or equal to a threshold based on the current access device list, the real-time scene mode information is determined to be a personal exclusive mode.

3. The method according to claim 1, characterized in that, Emotional state information obtained from real-time analysis of user facial images includes: Capture a video stream containing the user's face using a camera; Facial feature frames are extracted from the video stream and input into a pre-trained emotion recognition convolutional neural network model; Obtain the output of the convolutional neural network model, the output including multiple emotional states such as pleasure, calmness, fatigue, and irritability, and the confidence level corresponding to each emotional state; The emotional state with the highest confidence level is determined as the emotional state information.

4. The method according to claim 1, characterized in that, In response to a combined function setting instruction for the target media content, a composite operation protocol is generated for one-click triggering of a preset function sequence, including: Parse the combined function setting instructions to extract multiple discrete function operations and their corresponding parameters; According to the execution order of the discrete functional operations, each functional operation and its corresponding parameters are encoded into a structured data object. The data object includes an operation sequence list, and each item in the list is defined with a functional identifier and its parameter field. The structured data object is encapsulated into the composite operation protocol.

5. The method according to claim 1, characterized in that, Synchronizing the smart collection unit to multiple terminal devices associated with the current user account includes: When the smart TV and the multiple terminal devices have established a shared data ledger, corresponding data transaction records are generated; The data transaction records are distributed to multiple terminal devices associated with the current user account for consensus verification. The consensus verification includes verifying the digital signature of the data transaction records and verifying that the order of the data transaction records in the shared data ledger conforms to a preset rule. After receiving the consensus verification result from the associated terminal device, the data transaction record is appended to the shared data ledger of the smart TV, and the associated multiple terminal devices are instructed to append synchronously to achieve state synchronization of the smart collection unit.

6. The method according to claim 1, characterized in that, After synchronizing the smart collection unit to multiple terminal devices associated with the current user account, the method further includes: Obtain the current real-time context information; The real-time context information is matched and calculated with the multi-dimensional context information associated with the intelligent collection unit; Based on the matching calculation results, the multiple smart collection units that are currently available are sorted.

7. The method according to any one of claims 1 to 6, characterized in that, The method also includes: In response to a user's command to invoke the smart collection unit, the composite operation protocol bound to the smart collection unit is obtained; The composite operation protocol is parsed to obtain the preset function sequence; Control the smart TV to execute the preset function sequence to trigger the playback of target media content and simultaneously execute multiple function operations encapsulated in the preset function sequence.

8. A favorites generation device for a smart TV, comprising a processor and a memory storing program instructions, characterized in that, The processor is configured to execute, when running the program instructions, the method for generating favorites for a smart TV as described in any one of claims 1 to 7.

9. A smart TV, characterized in that, include: The smart TV itself; The favorites generation device for a smart TV as described in claim 8 is installed on the smart TV body.

10. A computer-readable storage medium storing program instructions, characterized in that, When the program instructions are executed, they cause the computer to perform the bookmark generation method for smart TVs as described in any one of claims 1 to 7.

Images