Information processing method and electronic equipment

Abstract

The invention provides an information processing method and electronic equipment, and is applied to the technical field of artificial intelligence. The information processing method comprises the steps of obtaining first input; analyzing the first input, and displaying at least one object based on the first parameter in response to the condition that the first input meets the condition; the first parameter and the at least one object are related to the first input; obtaining a second input; the second input comprises an operation acting on at least one object; analyzing the second input to obtain a second parameter, so that the first model generates target content at least based on the second parameter and the first input; the second parameter is a supplementary parameter of the first parameter.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of artificial intelligence, and in particular, to an information processing method and an electronic device. BACKGROUND

[0002] With the popularization of artificial intelligence technology, large language models have been widely applied in content creation, figure generation, and intelligent design, etc. Currently, users mainly interact with text instructions, and it is difficult to accurately understand the actual meaning of non-quantitative descriptions such as "slightly apart" and "a little bigger" when describing spatial relationships.

[0003] The existing solutions lack visual analysis and interactive adjustment mechanisms for ambiguous semantics, resulting in deviations between the generated results and user expectations, and multiple text corrections are required to achieve the desired effect, which seriously affects the efficiency of human-computer interaction and user experience. SUMMARY

[0004] Therefore, the present disclosure provides an information processing method and an electronic device.

[0005] According to a first aspect of the present disclosure, an information processing method is provided, the method comprising: obtaining a first input; analyzing the first input, and in response to the first input satisfying a condition, displaying at least one object based on a first parameter; the first parameter and the at least one object are related to the first input; obtaining a second input; the second input includes an operation acting on the at least one object; analyzing the second input to obtain a second parameter, so that a first model generates target content based on at least the second parameter and the first input; the second parameter is a complementary parameter of the first parameter.

[0006] According to an embodiment of the present disclosure, in response to the first input containing a non-quantitative description related to a visual space and the at least one object, it is determined that the first input satisfies the condition; the non-quantitative description is for the at least one object.

[0007] According to an embodiment of the present disclosure, displaying the at least one object based on the first parameter includes any one of the following: matching an exemplary object of the at least one object from a preset object library, and rendering the exemplary object according to the first parameter; inputting the first parameter into a second model to generate at least one personified object corresponding to the first input, and rendering the personified object according to the first parameter.

[0008] According to an embodiment of the present disclosure, if an adjustment operation for the rendered personified object is obtained, a first prompt word is determined based on the adjustment operation; the first prompt word is used to guide the first model to regenerate the corresponding personified object when generating the target content; if the adjustment operation for the rendered personified object is not obtained, a second prompt word is determined based on the adjustment operation; the second prompt word is used to guide the first model to directly use the already generated personified object for partial generation when generating the target content.

[0009] According to an embodiment of the present disclosure, the first input is subjected to recognition processing based on a third model to obtain structured data; or the first input is subjected to text generation processing based on a fourth model to obtain structured data, the structured data at least including a relation subject, a relation type and degree modification information; the relation subject is mapped to at least one object; and the relation type and the degree modification information are mapped to a first parameter.

[0010] According to an embodiment of the present disclosure, the first input is subjected to word segmentation and part-of-speech tagging by a third model to recognize subject information and modification information of the first input; and the third model is used to analyze the dependency relationship between the subject information and the modification information to generate structured data corresponding to the first input.

[0011] According to an embodiment of the present disclosure, the first input is filled into a corresponding placeholder in a preset prompt word template to obtain a third prompt word; and the third prompt word is input into a fourth model to generate structured data corresponding to the first input.

[0012] According to an embodiment of the present disclosure, in response to the at least one object being a single object, at least one of the color, size, transparency or shape of the exemplary object is rendered according to the display attribute parameter of the single object; and in response to the at least one object being a plurality of objects, the layout positions of the objects are determined based on the relative position parameter; and at least one of the relative distance, overlapping relationship, arrangement order of the plurality of exemplary objects and the color, size, transparency or shape of each object is rendered according to the layout positions and the display attribute parameters of the objects.

[0013] According to an embodiment of the present disclosure, an adjustment instruction for the display attribute parameter of the single object is received; or an adjustment instruction for the relative position parameter between the plurality of objects in the at least one object or the display attribute parameter of each object is received.

[0014] A second aspect of the present disclosure provides an electronic device, comprising: an input module configured to obtain a first input and a second input; a memory configured to store computer instructions and a first model; and a processor configured to load the computer instructions to perform: parsing the first input, and in response to the first input satisfying a condition, displaying at least one object based on a first parameter; and parsing the second input to obtain a second parameter; and the processor is further configured to load the first model to perform generating target content based on at least the second parameter and the first input.

[0015] A third aspect of the present disclosure provides a computer readable storage medium having stored thereon computer programs or instructions, which, when executed by a processor, implement the steps of the above method.

[0016] It is to be understood that the description of the contents of this section is not intended to identify key or essential features of embodiments of the disclosure, nor is it used to limit the scope of the disclosure. Other features of the disclosure will become readily apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0017] The above and other objects, features and advantages of the present disclosure will become more apparent from the following description when taken in conjunction with the accompanying drawings, in which:

[0018] Figure 1 A diagram illustrating an application scenario of an information processing method and an electronic device according to embodiments of the present disclosure is schematically shown;

[0019] Figure 2 A flowchart of an information processing method according to embodiments of the present disclosure is schematically shown;

[0020] Figure 3A A diagram schematically showing displaying an exemplary object according to embodiments of the present disclosure is shown;

[0021] Figure 3B A diagram schematically showing displaying an iconic object according to embodiments of the present disclosure is shown;

[0022] Figure 4A A diagram schematically showing displaying a single exemplary object according to embodiments of the present disclosure is shown;

[0023] Figure 4B A diagram schematically showing displaying a plurality of exemplary objects according to embodiments of the present disclosure is shown;

[0024] Figure 5 A diagram schematically showing displaying a plurality of exemplary objects after multiple inputs according to embodiments of the present disclosure is shown;

[0025] Figure 6 A block diagram of an electronic device according to embodiments of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0026] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely examples, and are not intended to limit the scope of the present disclosure. In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of embodiments of the present disclosure. However, it will be apparent to one skilled in the art that one or more embodiments can be practiced without these specific details. In other instances, well-known structures and techniques have been omitted in order to avoid obscuring the concepts of the present disclosure.

[0027] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. As used herein, the terms "comprises", "comprising", "includes", "including" and the like are specifically intended to be open-ended and to mean that other features, steps, operations, and / or components can be added.

[0028] All terms used herein including technical and scientific terms have the meanings commonly understood by one of ordinary skill in the art unless otherwise defined. It should be noted that the terms used herein are defined as having meanings that are consistent with the context of the specification in which they are used, and should not be interpreted in an overly idealized or overly formal way.

[0029] In the case of using expressions similar to "at least one of A, B, and C, etc.", it should generally be interpreted to include any of them alone, any combination of two or more of them, and the like in a manner that makes sense in the context of the specification in which it is used.

[0030] Before introducing the technical solutions provided by the embodiments of the disclosure, the related technologies involved in the disclosure are described.

[0031] In the related art, taking an intelligent dialogue as an example, the existing instruction processing mechanism, for example, can be that when receiving a user's spatial relationship description containing multiple objects, semantic analysis and content generation are directly performed based on internal knowledge of an algorithm, and information processing involves identification and execution of object attributes, spatial positions, and the like parameters, wherein the spatial relationship description contains feature information of multiple interactive objects, which is collectively referred to as a spatial instruction for brevity.

[0032] The existing spatial instruction processing mechanism, for example, can be that after a user instruction is received in a text input box, key elements are extracted through natural language processing technology and directly transmitted to a generation model; if a fuzzy description is contained, only approximate processing is performed by relying on preset parameters. However, when the spatial relationship is relatively complex, the user needs to make multiple text corrections to achieve the expected effect, which may cause dissatisfaction with the generated result and affect the creation efficiency.

[0033] As can be seen, the analysis of the spatial relationship in the existing instruction processing mechanism is implicit. When the user tries to accurately express the spatial intention, additional time is often spent on language modification, which may cause interruption and repetition of the creation process.

[0034] In view of this, the embodiments of the present disclosure provide an information processing method applied to an intelligent dialogue system, the method comprising: in response to obtaining a first input of a user; parsing the first input, and in response to the first input satisfying a preset condition, displaying at least one object based on a first parameter related to the first input; obtaining a second input generated by the user operating the at least one object; parsing the second input to obtain a second parameter, the second parameter being used as a supplement to the first parameter to guide a generation model to generate target content based on the first input and the second parameter.

[0035] Embodiments of the present disclosure can introduce a visual interaction mechanism of objects in the instruction processing process, which not only facilitates the user to intuitively confirm and accurately adjust the ambiguous intention, but also avoids the understanding deviation and multiple corrections caused by traditional pure text interaction. This interaction mechanism based on parameter supplement makes the user intention be more accurately captured and executed, thereby improving the efficiency and quality of content generation while improving the user experience.

[0036] The application scenarios of the embodiments of the present disclosure can include but are not limited to: intelligent painting, automatic presentation document making, flowchart generation, and other visual content creation platforms. It is particularly suitable for creative design scenarios, technical drawing scenarios, and multimedia content production scenarios that require accurate control of object spatial layout. In specific implementation scenarios, the method can also be applied to office software, teaching tools, and other intelligent applications integrated with intelligent generation functions.

[0037] Figure 1 An application scenario diagram of the information processing method and the electronic device according to the embodiments of the present disclosure is schematically shown.

[0038] As shown in Figure 1 The application scenario 100 according to the embodiments can include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 can include various connection types, such as wired, wireless communication links, or optical fiber cables, etc.

[0039] The user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. The client application installed on the terminal device can support user input instructions and receive the processing results returned by the server.

[0040] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with display screens and supporting graphical interface interaction, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like. The display screens of these devices are used to display at least one object generated based on the first parameters and receive user operation input on these objects.

[0041] The server 105 can be a server providing an intelligent content generation service, for example, a background server processing instructions submitted by a user through a terminal device (only as an example). The server 105 is configured to perform the following processing: parsing the received first input, when it is determined that the first input meets a preset condition, generating visual information based on the first parameters related to the first input and returning the visual information to the terminal device for display; receiving the second input uploaded by the terminal device, parsing to obtain the second parameters, and inputting the first input and the second parameters into the generation model together to generate target content.

[0042] It should be noted that the information processing method provided by the embodiments of the present disclosure can generally be executed by the server 105. Accordingly, the information processing apparatus provided by the embodiments of the present disclosure can generally be arranged in the server 105. The information processing method provided by the embodiments of the present disclosure can also be executed by the terminal device, or cooperatively executed by the terminal device and the server.

[0043] It should be understood that Figure 1 The number of terminal devices, networks, and servers in the above application scenario is only illustrative. According to the needs of implementation, there can be any number of terminal devices, networks, and servers.

[0044] In the above application scenario, when the user inputs the first input containing a vague description through the terminal device, the matching degree of the output result of the generation model and the user's intended intent can be effectively improved through the object visualization and parameter supplement mechanism.

[0045] Figure 2 A flowchart of an information processing method according to an embodiment of the present disclosure is schematically shown.

[0046] As Figure 2 indicated, the information processing method of this embodiment can be applied to an intelligent dialogue system, and the information processing method includes operations S210-S240.

[0047] In operation S210, the first input is obtained.

[0048] In the embodiments of the present disclosure, the first input refers to data carrying user intent. The data form of the first input includes but is not limited to text, voice or a combination of text and voice. For example, the user can input text data through a keyboard, or input voice through a microphone, and the voice data can be converted into text data through a voice recognition service. The text data and language data are taken as the first input for subsequent processing.

[0049] In the embodiments of the present disclosure, the first input can be a complete instruction of a single input, or an instruction sequence composed of multiple inputs. When the first input is an instruction sequence, the instruction sequence is semantically integrated to form a unified instruction containing a complete spatial relationship description for processing.

[0050] It should be noted that the first input serves as the starting point of the subsequent processing flow, providing an information basis for subsequent structured analysis and visual reconstruction, and ensuring that the spatial intent expressed by the user in natural language can be accurately captured and converted into an interactive visual representation.

[0051] In operation S220, the first input is parsed, and at least one object is displayed based on the first parameter in response to the first input satisfying a condition; the first parameter and the at least one object are related to the first input.

[0052] In the embodiments of the present disclosure, parsing refers to the process of semantic analysis and information extraction of the first input. The condition is determined based on the content feature recognition of the first input. When the first input contains a specific type of description, it is determined that the condition is satisfied. The object refers to a visual element presented in the display interface. The first parameter refers to a configuration parameter used to control the display characteristics of the object.

[0053] It should be noted that the characteristics of the at least one object are determined by the content of the first input, and the value of the first parameter is determined based on the parsing result of the first input. This correlation ensures the correspondence between the visual presentation and the user input, so that the displayed object and its configuration can reflect the content features of the first input.

[0054] In operation S230, a second input is obtained; the second input includes an operation on the at least one object.

[0055] In the embodiments of the present disclosure, the second input refers to an operation instruction executed by the user on the at least one object displayed in the display interface through the input device, and the second input reflects the adjustment intent expressed by the user on the configuration of the at least one object. The data generated by the operation instruction will be used as the input basis for the subsequent generation process. By obtaining the second input, the user can interact with the initial display visual result, thereby realizing the optimization and correction of the initial configuration.

[0056] In operation S240, the second input is parsed to obtain a second parameter, so that the first model generates target content based on at least the second parameter and the first input; the second parameter is a supplementary parameter of the first parameter.

[0057] In the embodiments of the present disclosure, parsing the second input to obtain the second parameter refers to a processing process of data extraction and parameterization conversion of the user's interactive operation in the visualization interface. Specifically, by capturing the operation trajectory and operation type of the user on the displayed object, the operation trajectory and operation type are converted into a set of quantitative parameters.

[0058] The second parameter as a supplementary parameter of the first parameter is a correction and improvement of the initial space configuration. The first parameter provides an initial space relationship framework based on text understanding, and the second parameter records the adjustment of the initial space relationship framework based on visual feedback by the user. The supplementary relationship between the second parameter and the first parameter maintains the coherence of the original design intention and integrates the user's interactive correction.

[0059] In the embodiments of the present disclosure, by converting the non-quantitative description in the user input into a visualization object and generating a supplementary parameter based on user interaction, the content generation model can accurately understand the user's space relationship intention. A conversion mechanism from fuzzy language description to accurate space parameters is established, and through parameter supplementing and fusion processing, the consistency between the generated content and the user's expectation is significantly improved.

[0060] The above operation S220 can further include: in response to the first input containing a non-quantitative description related to a visual space and at least one object, determining that the first input satisfies the condition; the non-quantitative description is for the at least one object.

[0061] In the embodiments of the present disclosure, the first input satisfying the condition can be based on joint analysis of semantic elements in the first input. There is description content related to visual space characteristics in the semantic elements in the first input. The visual space characteristics refer to the space attributes related to the visual presentation of the object, including but not limited to the relative position relationship between objects, the size ratio relationship of the object, and the spatial hierarchy relationship of the object. The non-quantitative description refers to a description method lacking accurate numerical indicators, which is characterized by using degree adverbs for modification, using comparative degrees for expression, and using fuzzy quantifiers.

[0062] For example, when the first input is "place the picture on the left side of the text and keep a proper distance", where "picture" and "text" constitute at least one object, "place on the left side" and "proper distance" constitute a visual space non-quantitative description for the object, and the description directly acts on the object, it is determined that the condition is satisfied.

[0063] When the first input is "adjust the button size to be slightly larger than the icon", where "button" and "icon" constitute at least one object, "slightly larger" constitutes a non-quantitative description of the size ratio of the object, and the description is directly related to the object, so it is also determined that the condition is met.

[0064] In the embodiments of the present disclosure, the first input meeting the condition can also be based on that there is an explicit described object in the statement of the first input, and the non-quantitative description is directly related to the described object in semantics. The object has an explicit non-quantitative description in the statement of the first input, and there is a direct modification relationship with the object in the grammatical structure, and the spatial change expressed by the non-quantitative description directly acts on the object.

[0065] For example, when the first input is "the title and the content are slightly separated", "slightly separated" constitutes a visual spatial non-quantitative description, "title" and "content" constitute an explicit described object, and "slightly separated" is directly related to and acts on "title" and "content" in semantics, so the first input meets the condition.

[0066] In the embodiments of the present disclosure, the non-quantitative spatial relationship description between the objects in the first input is recognized, and the visualization process is triggered when there is a direct semantic association with a specific object. The fuzzy spatial expression in natural language is automatically converted into a visualization generation instruction, a conversion mechanism from language description to spatial parameters is established, the problem of difficulty in understanding non-quantitative spatial relationship by traditional interactive systems is effectively solved, and the accuracy and efficiency of human-computer interaction are significantly improved.

[0067] The displaying at least one object based on the first parameter in the operation S220 can further include any one of the following: matching an example object of the at least one object from a preset object library according to the first parameter, and rendering the example object according to the first parameter; inputting the first parameter into a second model to generate at least one personified object corresponding to the first input, and rendering the personified object according to the first parameter.

[0068] In the embodiments of the present disclosure, according to the object identification features parsed from the first input, similarity matching and retrieval are performed in a preset object library to obtain example object data closest to the user description. The object library stores object templates of various types and attribute information thereof.

[0069] After obtaining the example object data, the example object is parameterized and rendered based on the spatial configuration information contained in the first parameter, including but not limited to size ratio, position coordinates, rotation angle, and other display attributes. The rendering process adjusts the geometric transformation parameters and visual style parameters of the object, so that the example object meets the spatial constraint conditions defined by the first parameter, and completes the visualization presentation conforming to the user's intention on the display interface.

[0070] Figure 3A A schematic diagram of displaying an example object is illustratively shown.

[0071] As Figure 3A shown, after the user inputs "generate a cuboid and a cylinder, the cylinder is a little bit to the right of the cuboid" in the input interface, the following first parameters are obtained through parsing processing: the relationship subject is the cuboid 301A, the cylinder 301A, the position relationship degree modification is "a little bit", the spatial configuration is the relative position "to the right", and the interval degree is slight.

[0072] Based on the first parameters, the following processing is performed: matching the corresponding example object from the pre-set object library; obtaining the standard cuboid model data and obtaining the standard cylinder model data; parameterizing rendering according to the first parameters; based on the relative position parameter, placing the cylinder to the right of the cuboid; based on the degree modification "a little bit", setting the interval between the two to be 10% of the default interval; keeping the default size and material properties of the two objects; completing the visual presentation in the display interface; displaying the cuboid figure and the cylinder figure at a proper distance to the right of the cuboid; and keeping the relative position relationship between the two objects in the visual aspect.

[0073] In the embodiment of the present disclosure, the spatial configuration information contained in the first parameters is converted into input instructions understandable by the generation model, which contains spatial constraint conditions and style requirements for object generation. The generation model performs deep processing on the input instructions, and outputs object data that meets the spatial constraint conditions. The generation model is trained to generate object instances with corresponding visual features according to the spatial configuration information.

[0074] Based on the first parameters, the generated visualized object is fine-tuned and rendered, including accurate calibration of spatial position and adaptive adjustment of display attributes, to ensure that the generated object not only meets the creative requirements, but also strictly follows the spatial relationship defined by the first parameters, and completes the visual presentation in the display interface that meets the user's description and has visual expressiveness.

[0075] Figure 3B A schematic diagram of displaying a visualized object is illustratively shown.

[0076] As Figure 3B shown, after the user inputs "generate a cuboid and a cylinder, the cylinder is a little bit to the right of the cuboid" in the input interface, the following first parameters are obtained through parsing processing: the relationship subject is the cuboid 301A, the cylinder 301A, the position relationship degree modification is "a little bit", the spatial configuration is the relative position "to the right", and the interval degree is slight.

[0077] The two iconic objects are rendered based on the first parameter, the relative positions are calibrated and the interval is adjusted to the reference distance, and the iconic objects are presented on the display interface.

[0078] In the embodiments of the present disclosure, based on the object library matching and the model generation object visualization scheme, the automatic conversion from the natural language description to the visualized elements is realized by directly mapping the spatial parameters to the rendering instructions or the generation constraints of the generated model, the non-quantitative spatial relationship is converted into the accurate visualized layout through the parameterized rendering mechanism, the problem that the user needs to manually create and adjust the graphical elements is effectively solved, and the efficiency and accuracy of the interface design are significantly improved.

[0079] In the embodiments of the present disclosure, rendering the exemplary object according to the first parameter can further include: in response to the at least one object being a single object, rendering at least one of a color, a size, a transparency or a shape of the exemplary object according to a display attribute parameter of the single object.

[0080] In the embodiments of the present disclosure, rendering the exemplary object according to the first parameter includes a configuration process of the object display attribute. When the at least one object is a single object, one or more visual features of the exemplary object are adaptively rendered according to the display attribute parameter contained in the first parameter. The display attribute parameter includes but is not limited to at least one of a color value, a size ratio, a transparency level or a geometric shape, and the visualized presentation of the exemplary object is realized by adjusting the corresponding graphical rendering parameter.

[0081] Figure 4A A schematic diagram of displaying a single exemplary object is shown.

[0082] As Figure 4A shown, when the user inputs "adjust the circle to a slightly larger green circle", a circle is matched from the preset object library as the exemplary object, and the display attribute parameter in the first parameter is used. Based on the size ratio parameter, the rendering size of the circle is increased; based on the color value parameter, the circle is rendered as green; and the rendering and display of the single exemplary object are completed.

[0083] In the embodiments of the present disclosure, rendering the exemplary object according to the first parameter can further include: in response to the at least one object being a plurality of objects, determining layout positions of the objects based on a relative position parameter; and rendering at least one of a relative distance, an overlapping relationship, an arrangement order of the plurality of exemplary objects, and at least one of a color, a size, a transparency or a shape of each object according to the layout positions and display attribute parameters of the objects.

[0084] In the embodiments of the present disclosure, when rendering involves multiple objects, the position relationship of each object in the layout space is determined based on the relative position parameter in the first parameter, and the overall rendering is completed in combination with the display attribute parameter of each object. The relative position parameter defines the spatial topological relationship between objects, including at least one of a relative distance vector, an overlap priority, and an arrangement order; and the display attribute parameter controls the visual features of each object itself, including at least one of a color space value, a size ratio, a transparency coefficient, or a geometric shape type. By comprehensively processing the spatial layout parameter and the object attribute parameter, the spatial relationship and individual features between multiple objects are cooperatively rendered.

[0085] Figure 4B A schematic diagram of displaying multiple exemplary objects is shown.

[0086] As shown in Figure 4B For example, when the user inputs "make a page of presentation, title and content slightly apart", the following rendering operations are performed: based on the relative position parameter, the vertical layout relationship of the title box and the content box is established, the title box is positioned above the content box, and the vertical spacing conforming to the visual comfort is set according to the degree modification "slightly"; according to the display attribute parameter, the title box is rendered as a standard size rectangular area conforming to the title level, and the content box is rendered as a standard size rectangular area suitable for text arrangement; the alignment relationship of the two rectangular areas in the vertical direction is maintained, ensuring that the title box and the content box maintain the same horizontal starting position and width value, forming a page layout conforming to the presentation specification.

[0087] In the embodiments of the present disclosure, based on the object visualization scheme of parameterized rendering, the display attribute parameter and the relative position parameter are directly mapped into graphic rendering instructions, realizing accurate visual control of a single object or multiple objects. By establishing the correspondence between the parameters and the visual features, the attributes of the objects such as color, size, and transparency, and the spatial relationship between the objects such as relative position and arrangement order can be automatically processed, effectively solving the problem of low efficiency of traditional manual adjustment, and significantly improving the accuracy and operation efficiency of visualization generation.

[0088] The obtaining of the second input in the above operation S240 can further include: receiving an adjustment instruction for the display attribute parameter of a single object; or, receiving an adjustment instruction for the relative position parameter between multiple objects in the at least one object or the display attribute parameter of each object.

[0089] In the embodiments of the present disclosure, obtaining the second input includes receiving an adjustment instruction of the displayed object by the user. The adjustment instruction is divided into two types: the first type is for a single object, involving modification of the display attribute parameter thereof; the second type is for multiple objects, involving adjustment of the relative position parameter between the objects or the display attribute parameter of each object. These adjustment instructions are obtained by capturing the interactive operation of the user on the graphical interface and converted into corresponding parameter modification information.

[0090] For example, after the interface displays the title box and the content box, the user can generate the second input by the following operations: dragging the title box to change the position of the title box, for example, adjustment of the display attribute parameter of a single object; simultaneously selecting the title box and the content box and adjusting the distance therebetween, for example, adjustment of the relative position parameter between multiple objects; respectively modifying the color of the title box and the size of the content box, for example, adjustment of the display attribute parameter of each object. These adjustment operations will be captured and converted into corresponding second input parameters for subsequent content generation processing.

[0091] After displaying the at least one object based on the first parameter in the operation S220, the method can further include, after displaying the visualized object, if an adjustment operation for the rendered visualized object is obtained, determining a first prompt word based on the adjustment operation; the first prompt word is used to guide the first model to regenerate the corresponding visualized object when generating the target content.

[0092] In the embodiments of the present disclosure, after displaying the at least one visualized object based on the first parameter, if an adjustment operation performed by the user on the visualized object is detected, a first prompt word is generated based on the adjustment operation. The first prompt word contains description information of the modification of the characteristics of the visualized object, which is used to guide the first model to regenerate the corresponding visualized object according to the adjusted characteristics when generating the target content.

[0093] For example, after displaying the initial visualized object according to the user input "generate a circular button", if the user adjusts the circular shape into an elliptical shape through a dragging operation, a first prompt word "modify the circular shape into an elliptical shape" is generated based on the adjustment operation. The first prompt word will be input into the first model together with the original input, guiding the model to output a button image with an elliptical shape feature in the finally generated target content.

[0094] After displaying the at least one object based on the first parameter in the operation S220, the method can further include, after displaying the visualized object, if an adjustment operation for the rendered visualized object is obtained, determining a first prompt word based on the adjustment operation; the first prompt word is used to guide the first model to regenerate the corresponding visualized object when generating the target content.

[0095] In the embodiments of the present disclosure, after displaying the at least one visualized object based on the first parameter, if no adjustment operation performed by the user on the visualized object is detected within a preset time threshold, a second prompt word is automatically generated. The second prompt word contains confirmation information of the generated characteristics of the visualized object, which is used to guide the first model to directly reference the rendered visualized object as a generation basis when generating the target content.

[0096] For example, after displaying the corresponding visualized object according to the user input "generate a blue rectangular icon", if the user does not perform any adjustment operation within the set time, a second prompt word "keep the current blue rectangular style" is automatically generated. The second prompt word will be input into the first model together with the original input, guiding the model to directly use the rendered blue rectangular icon as a component in the finally generated target content.

[0097] In the embodiments of the present disclosure, by monitoring the adjustment operation of the user on the visualized object, the corresponding prompt word is intelligently generated: when the adjustment is detected, the first prompt word for guiding the content regeneration is generated, and when the adjustment is not detected, the second prompt word for guiding the direct reference to the existing visualized object is generated. The accurate conversion of the user interaction intention to the generation instruction is realized, which not only ensures the high consistency between the generated content and the user's expectation, but also avoids unnecessary repeated generation, significantly improving the content generation efficiency and quality.

[0098] The parsing the first input in the operation S220 can further include: performing identification processing on the first input based on a third model to obtain structured data; or performing text generation processing on the first input based on a fourth model to obtain structured data, the structured data at least including a relationship subject, a relationship type, and degree modification information; mapping the relationship subject to at least one object; and mapping the relationship type and the degree modification information to the first parameter.

[0099] In the embodiments of the present disclosure, the parsing process realizes the extraction of structured data through two optional technical paths. The first technical path adopts a language analysis model to perform syntax analysis and semantic recognition on the input text, and constructs a structured data representation; the second path adopts a text generation model, and guides the text generation model to directly output structured data in a standard format through a templated prompt word. The structured data includes three elements: a relationship subject, a relationship type, and degree modification information, wherein the relationship subject represents an operation object entity, the relationship type describes the spatial interaction characteristics between the subjects, and the degree modification information quantifies the non-quantitative description of the relationship strength.

[0100] For example, when the user inputs "the title should be slightly separated from the content", "title" and "content" are identified as relationship subjects, "separated" is identified as a relationship type, and "slightly" is identified as degree modification information, and structured data containing these elements is generated.

[0101] In the embodiments of the present disclosure, the natural language input is converted into structured data containing the relationship subject, the relationship type and the degree modification through syntax analysis by a language analysis model or direct generation by a text generation model. The automatic conversion from vague description to accurate parameters is realized, and the accurate mapping between natural language and visual parameters is established.

[0102] Figure 5 A schematic diagram illustrating that multiple exemplary objects are displayed after multiple inputs of the embodiments of the present disclosure is shown.

[0103] As shown in Figure 5 for example, after the user inputs "draw an algorithm flowchart, the rectangle in the diagram is slightly larger than the diamond box; then make a page of document for this flowchart, the title needs to be slightly smaller than the flowchart, and the content occupies an area similar to the flowchart", the following rendering operations are performed: first, the flowchart shape elements and the document structure elements are identified, according to the degree modification "slightly larger", the rectangle box is rendered to be obviously larger than the diamond box; based on the indication "slightly smaller", the title area is set to be obviously smaller than the size specification of the flowchart; according to the requirement "similar area", the content area is maintained to be similar to the flowchart in the occupied area.

[0104] In terms of overall layout, the rectangle box and the diamond box are combined and arranged according to the flowchart specification, and the centers of the two are aligned and the spacing is coordinated. The title area is placed in the center of the top of the page, the flowchart is placed in the main area of the page, and the content area is arranged side by side with the flowchart to maintain the visual balance and proportional coordination between the areas.

[0105] Each element is strictly rendered and presented according to the recognized spatial relationship and size requirement, the rectangle box and the diamond box use differentiated filling colors to realize visual differentiation, the title area uses an emphasized font, and the content area maintains a standard text box style, finally forming a complete document layout preview that meets the user's requirements.

[0106] In the embodiments of the present disclosure, the third model performs word segmentation and part-of-speech tagging on the first input to identify the subject information and the modification information of the first input; the third model analyzes the dependency relationship between the subject information and the modification information to generate the structured data corresponding to the first input.

[0107] In the embodiments of the present disclosure, the third model can be a language analysis model constructed based on traditional natural language processing technology, the third model forms a complete language understanding pipeline by integrating core components such as a word segmenter, a part-of-speech tagger and a dependency syntax analyzer, and the first input can be an input text.

[0108] The third model performs lexical analysis and grammatical role labeling on the input text through linguistic analysis techniques, identifies the core semantic components and their modification relationships in the sentence, establishes a grammatical association network between words through dependency relation analysis, locates the core predicate describing the spatial relationship and the core predicate associated subject, extracts subject, relationship, and modification triples based on grammatical structure features, and generates machine-readable structured data.

[0109] For example, when processing the input sentence "Make a page of presentation, title and content to be slightly separated", "separate" is identified as the core predicate, "title" as the action subject, "content" as the parallel subject, and "slightly" and "a little" as the degree modifier through syntactic analysis; based on the pre-defined semantic feature library, it is confirmed that "separate" belongs to the spatial relationship verb and "slightly" belongs to the degree modifier; the relationship subject is "title" and "content", the relationship type is distance relationship, and the degree of modification is slight level; according to the degree of modification mapping rule, "slightly" is quantified as a reference distance parameter of about 10%, and structured data containing these elements is generated.

[0110] In the embodiments of the present disclosure, the third model performs word segmentation, part-of-speech tagging, and dependency relation analysis on the first input, automatically extracts the subject information and modification information in the sentence, and generates structured data containing complete semantic elements. It realizes accurate conversion from natural language to structured data, effectively solving the technical problem of non-quantitative spatial relationship difficult to quantify and analyze.

[0111] In the embodiments of the present disclosure, the first input is filled into the corresponding placeholders in the preset prompt template to obtain a third prompt; the third prompt is input into the fourth model to generate structured data corresponding to the first input.

[0112] In the embodiments of the present disclosure, the preset prompt template includes instruction description, format definition, and example description for guiding the model to output standardized data. By filling the user input into a specific position in the preset prompt template, a complete prompt word that meets the processing requirements of the fourth model is constructed. The fourth model directly generates structured data output that meets the requirements according to the task description and format constraints in the prompt word.

[0113] It should be noted that the preset prompt template explicitly defines the professional analysis role of the fourth model in the task; the specific preset prompt template includes task description, processing rule, and output specification. Among them, the task description clearly describes the instruction analysis task to be completed; the processing rule lists the core principles of information extraction in detail, including the identification requirements for non-quantitative description features; the output specification strictly specifies the output of structured data in a standardized format.

[0114] For example, when the user inputs "title and content should be slightly separated", the sentence is filled into the input position of the template, and the complete prompt word formed will guide the model to output standardized structured data containing the identified relationship subjects "title" and "content", the relationship type "distance", and the degree modifier "slightly".

[0115] In the embodiments of the present disclosure, the user input is converted into a structured query instruction through a preset prompt word template, and standardized structured data is generated by using a fourth model. The semantic ambiguity problem in natural language parsing is effectively solved, and the accuracy and consistency of relationship element extraction are ensured.

[0116] The parsing of the second input in the above operation S240 can further include a parameterization conversion process.

[0117] In the embodiments of the present disclosure, the parsing process includes but is not limited to coordinate transformation analysis, geometric transformation calculation, and spatial relationship reconstruction. The coordinate transformation analysis can be through capturing the starting coordinates and ending coordinates of the user's drag operation, the geometric transformation calculation can be according to the gesture trajectory of the user's zoom, rotation, etc. operation, calculating the size change ratio or rotation angle of the object, and the spatial relationship reconstruction can be according to the gesture trajectory of the user's zoom, rotation, etc. operation, calculating the size change ratio or rotation angle of the object.

[0118] The second parameter refers to a parameter set obtained by parsing the user interaction operation, and the parameter set includes but is not limited to a space correction parameter, a relationship adjustment parameter, and an attribute correction parameter. The space correction parameter can be a coordinate offset of the object position, a scaling ratio of the object size, and an angle value of the object rotation, the relationship adjustment parameter can be a relative distance change value between multiple objects, an overlapping area ratio, and a spatial hierarchy order, and the attribute correction parameter can be a color value, a transparency, and a modification amount of a shape style.

[0119] Based on the above information processing method, the present disclosure further provides an electronic device. The following will be described in combination with Figure 6 The electronic device will be described in detail.

[0120] Figure 6 A block diagram of an electronic device according to an embodiment of the present disclosure is schematically shown.

[0121] As Figure 6 shown, the electronic device 600 of this embodiment includes an input module 610, a memory 620, and a processor 630.

[0122] The input module 610 is configured to obtain the first input and the second input of the user.

[0123] The memory 620 is configured to store computer instructions and a first model.

[0124] The processor 630 is connected to the input module 610 and the memory 620, respectively, and is configured to:

[0125] load the computer instructions to perform the following operations: parse the first input, in response to the first input satisfying a preset condition, control display of at least one object based on a first parameter related to the first input; parse the second input to obtain a second parameter, the second parameter being a complementary parameter of the first parameter; and load the first model to generate target content based on at least the second parameter and the first input.

[0126] For example, the input module 610 can include, but is not limited to, a touch screen, a keyboard, a mouse, a voice input device, and the like, and a corresponding driver controller.

[0127] For example, the memory 620 can include a non-volatile computer readable storage medium for storing the computer instructions and the first model, including but not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0128] For example, the processor 630 can include a general-purpose microprocessor (such as a CPU), a graphics processor (GPU), an instruction set processor, and / or a related chipset, and / or a special-purpose microprocessor (such as an application-specific integrated circuit (ASIC)), and the like. The processor 630 can also include an on-board memory for cache purposes. The processor 630 can also be configured to output a display signal to a display device to control the display of the at least one object and the target content.

[0129] The present disclosure also provides a computer readable storage medium, which can be included in the device / apparatus described in the above embodiments, or can exist separately without being assembled into the device / apparatus. The above computer readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of the present disclosure. The computer readable storage medium can be a non-volatile computer readable storage medium, which can include, but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0130] Those skilled in the art can understand that the features described in various embodiments of the present disclosure can be combined and / or integrated in various combinations, even if such combinations or integrations are not explicitly described in the present disclosure. In particular, the features described in various embodiments of the present disclosure can be combined and / or integrated in various combinations without departing from the spirit and teachings of the present disclosure. All such combinations and / or integrations are within the scope of the present disclosure.

[0131] The above describes embodiments of the present disclosure. However, these embodiments are merely for illustrative purposes, and are not intended to limit the scope of the present disclosure. Although each embodiment is described above separately, this does not mean that the measures in various embodiments cannot be used advantageously in combination. Those skilled in the art can make various substitutions and modifications without departing from the scope of the present disclosure, and these substitutions and modifications should all fall within the scope of the present disclosure.

Claims

1. An information processing method, comprising: Get the first input; Parse the first input, and in response to the first input satisfying the condition, display at least one object based on the first parameter; The first parameter and the at least one object are related to the first input; Obtain a second input; the second input includes an operation performed on the at least one object; The second input is parsed to obtain the second parameter, so that the first model generates target content based at least on the second parameter and the first input; The second parameter is a supplementary parameter to the first parameter.

2. The method according to claim 1, wherein the response to the first input satisfying the condition includes: In response to the first input containing a non-quantitative description related to visual space and at least one object, it is determined that the first input satisfies the condition; The non-quantitative description refers to the at least one object.

3. The method according to claim 1, wherein displaying at least one object based on the first parameter includes any of the following: Match an exemplary object of the at least one object from a preset object library, and render the exemplary object according to the first parameter; The first parameter is input into the second model to generate at least one visualized object corresponding to the first input, and the visualized object is rendered according to the first parameter.

4. The method according to claim 3, further comprising: If an adjustment operation is obtained for the rendered visualization object, a first prompt word is determined based on the adjustment operation; The first prompt word is used to guide the first model to regenerate the corresponding visual object when generating the target content; If no adjustment operation is obtained for the rendered visualization object, a second prompt word is determined based on the adjustment operation; The second prompt word is used to guide the first model to directly use the already generated visualization object to partially generate the target content.

5. The method according to claim 1, wherein parsing the first input comprises: The first input is processed based on the third model to obtain structured data; Alternatively, the first input can be processed by text generation based on the fourth model to obtain structured data, which includes at least relation subjects, relation types, and degree modification information; Map the relation subject to the at least one object; The relationship type and degree modification information are mapped to the first parameter.

6. The method according to claim 5, wherein the step of performing recognition processing on the first input based on the third model to obtain structured data includes: The third model is used to segment words and tag parts of speech in the first input, thereby identifying the main information and modifier information of the first input. The third model analyzes the dependency relationship between the main information and the modification information to generate structured data corresponding to the first input.

7. The method according to claim 5, wherein the text generation processing of the first input based on the fourth model to obtain structured data includes: Fill the first input into the corresponding placeholder in the preset prompt word template to obtain the third prompt word; The third prompt word is input into the fourth model to generate structured data corresponding to the first input.

8. The method according to claim 3, wherein the first parameter includes display attribute parameters of a single object, or relative position parameters between multiple objects and / or display attribute parameters of each object; and rendering the exemplary object according to the first parameter includes: In response to the fact that the at least one object is a single object, at least one of the color, size, transparency, or shape of the exemplary object is rendered according to the display attribute parameters of the single object; In response to the fact that the at least one object is multiple objects, the layout position of each object is determined based on the relative position parameter; Based on the layout position and the display attribute parameters of each object, render the relative distance, overlap, arrangement order, and at least one of the color, size, transparency, or shape of the multiple exemplary objects.

9. The method according to claim 8, wherein obtaining the second input comprises: Receive instructions to adjust the display attribute parameters of the individual object; or, Receives adjustment instructions for the relative position parameters between multiple objects in the at least one object or for the display attribute parameters of each object.

10. An electronic device, comprising: The input module is used to obtain the first input and the second input; Memory, used to store computer instructions and the first model; A processor is configured to load the computer instructions to execute: parsing the first input, and in response to the first input satisfying a condition, displaying at least one object based on a first parameter; and parsing the second input to obtain a second parameter; The processor is further configured to load the first model to generate target content based at least on the second parameter and the first input.

Images