Clothing change identification method and system, electronic equipment and storage medium

Abstract

The invention discloses a clothes change identification method and system, electronic equipment and a storage medium. The clothes change identification method comprises the following steps: acquiring an image description model and a text similarity identification model; acquiring a clothes reference image and a to-be-identified clothes image; and through the image description model and the text similarity identification model, obtaining a clothes change identification result according to the clothes reference image and the to-be-identified clothes image. The method combines the image description model and the text similarity recognition model to automatically judge whether the clothes in the image change, can improve the recognition accuracy of the change of the clothes in the inner barrel of the washing machine, reduces the human input cost, and can be widely applied to the technical field of artificial intelligence.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, system, electronic device, and storage medium for identifying clothing changes. Background Art

[0002] In the laundry recognition scenario, the laundry content in a washing machine often remains unchanged when the image is captured. However, in terms of model recognition capabilities, images of different laundry content are more helpful for the final comprehensive recognition. Images of similar laundry content (similar types and minimal changes in the position of the laundry) are not very helpful for model training and testing. Furthermore, due to cost considerations, edge computing power is often very limited in actual washing machine laundry recognition.

[0003] Traditional image processing methods use pixel-by-pixel comparison or other image similarity metrics, such as PSNR and SSIM. These methods suffer from slow computational speed, low accuracy, significant environmental impact, and high computational volatility. For example, in a washing machine, a slight rotation of the inner drum barely changes the shape and position of the clothes, yet the resulting computational errors can be significant. Deep learning-based similarity comparison methods also face challenges due to their high dataset quality requirements and high labor costs. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a clothing change recognition method, system, electronic device and storage medium with higher accuracy and lower cost.

[0005] To achieve the above objectives, one aspect of an embodiment of the present application provides a method for identifying clothing changes, comprising the following steps:

[0006] Obtain image description model and text similarity recognition model;

[0007] Acquire a clothing reference image and an image of clothing to be identified;

[0008] A clothing change recognition result is obtained according to the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model.

[0009] In some embodiments, the clothing change recognition method further includes:

[0010] A clothing standard vocabulary is defined, wherein the clothing standard vocabulary includes clothing color, clothing material, clothing type, clothing position, and clothing status.

[0011] In some embodiments, acquiring the image description model and the text similarity recognition model specifically includes:

[0012] Acquire a sample data set, wherein the sample data set includes a plurality of continuously changing clothing image samples in various washing scenes;

[0013] Performing a text description on each clothing image sample according to the clothing standard vocabulary to obtain a description text sample corresponding to the clothing image sample;

[0014] A plurality of clothing image samples are input into a pre-built first image description model for training, and parameters of the trained first image description model are optimized according to the description text samples to obtain the image description model.

[0015] In some embodiments, acquiring the image description model and the text similarity recognition model specifically includes:

[0016] Performing similarity comparison on the description text samples that change continuously before and after to obtain a similarity comparison result;

[0017] A plurality of the description text samples are input into a preset classification model for training, and parameters of the trained classification model are optimized according to the similarity comparison result to obtain the text similarity recognition model.

[0018] In some embodiments, obtaining a clothing change recognition result based on the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model specifically includes:

[0019] Inputting the clothing reference image into the image description model to obtain a first description text;

[0020] Inputting the clothing image to be identified into the image description model to obtain a second description text;

[0021] The first description text and the second description text are input into the text similarity recognition model to obtain the clothing change recognition result.

[0022] In some embodiments, inputting the first description text and the second description text into the text similarity recognition model to obtain the clothing change recognition result specifically includes:

[0023] Concatenate the first description text and the second description text in a preset order to obtain a description text vector;

[0024] Inputting the description text vector into the text similarity recognition model for classification to obtain similarity categories, wherein the similarity categories include similarity and dissimilarity;

[0025] When the similarity category is similar, determining that the image of the clothing to be identified has not changed;

[0026] When the similarity category is dissimilar, it is determined that the to-be-identified clothing image has changed.

[0027] In some embodiments, inputting the first description text and the second description text into the text similarity recognition model to obtain the clothing change recognition result specifically includes:

[0028] Set similarity threshold;

[0029] Concatenate the first description text and the second description text in a preset order to obtain a description text vector;

[0030] Inputting the description text vector into the text similarity recognition model for grading to obtain a similarity level;

[0031] When the similarity level is equal to or greater than the similarity threshold, determining that the image of the clothing to be identified has not changed;

[0032] When the similarity level is less than the similarity threshold, it is determined that the to-be-recognized clothing image has changed.

[0033] To achieve the above objectives, another aspect of the present application provides a clothing change recognition system, comprising:

[0034] The first module is used to obtain an image description model and a text similarity recognition model;

[0035] The second module is used to obtain clothing reference images and clothing images to be identified;

[0036] The third module is configured to obtain a clothing change recognition result based on the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model.

[0037] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application proposes an electronic device, which includes a memory, a processor, a program stored on the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory. When the program is executed by the processor, the clothing change recognition method as described above is realized.

[0038] To achieve the above-mentioned purpose, another aspect of an embodiment of the present application proposes a storage medium, which is a computer-readable storage medium used for computer-readable storage. The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the clothing change identification method as described above.

[0039] The beneficial effects of the present invention are as follows: the clothing change recognition method, system, electronic device, and storage medium of the present invention first obtain an image description model and a text similarity recognition model, then obtain a clothing reference image and an image of clothing to be recognized. Finally, using the image description model and the text similarity recognition model, the clothing change recognition result is obtained based on the clothing reference image and the image of clothing to be recognized. The present invention combines the image description model and the text similarity recognition model to automatically determine whether the clothing in the image has changed, thereby improving the accuracy of clothing change recognition in the washing machine drum and reducing labor input costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following introduction is made to the drawings required for use in the embodiments of the present invention. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative work.

[0041] Figure 1 A flowchart of the steps of a clothing change recognition method provided by an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of the structure of an image description model provided by one embodiment of the present invention;

[0043] Figure 3 A schematic diagram of the steps for obtaining clothing change recognition results provided by an embodiment of the present invention;

[0044] Figure 4 A schematic structural diagram of a clothing change recognition system provided by an embodiment of the present invention;

[0045] Figure 5 A schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are merely examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the appended claims.

[0047] It will be understood that the terms "first", "second", etc. used in this application may be used herein to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0048] The terms "at least one", "plurality", "each", "any", etc. used in this application include "at least one", "two" or more, "plurality" or "each", "any" or "any one", "each" or "any one" as used herein.

[0049] In the laundry recognition scenario, the laundry content in a washing machine often remains unchanged when the image is captured. However, in terms of model recognition capabilities, images of different laundry content are more helpful for the final comprehensive recognition. Images of similar laundry content (similar types and minimal changes in the position of the laundry) are not very helpful for model training and testing. Furthermore, due to cost considerations, edge computing power is often very limited in actual washing machine laundry recognition.

[0050] Traditional image processing methods use pixel-by-pixel comparison or other image similarity metrics, such as PSNR and SSIM. These methods suffer from slow computational speed, low accuracy, significant environmental impact, and high computational volatility. For example, in a washing machine, a slight rotation of the inner drum barely changes the shape and position of the clothes, yet the resulting computational errors can be significant. Deep learning-based similarity comparison methods also face challenges due to their high dataset quality requirements and high labor costs.

[0051] To this end, embodiments of the present invention propose a clothing change recognition method. This method first acquires an image description model and a text similarity recognition model, then obtains a clothing reference image and an image of the clothing to be recognized. Finally, using the image description model and the text similarity recognition model, a clothing change recognition result is obtained based on the clothing reference image and the image to be recognized. By combining the image description model and the text similarity recognition model to automatically determine whether clothing in an image has changed, the present invention can improve the accuracy of clothing change recognition in a washing machine drum and reduce labor costs.

[0052] Reference Figure 1 , Figure 1 This is a flowchart of a method for identifying clothing changes according to an embodiment of the present invention. The method includes steps S101 to S103:

[0053] S101, obtaining an image description model and a text similarity recognition model;

[0054] As an optional embodiment, the clothing change recognition method further includes the following steps S1011:

[0055] S1011. Define a clothing standard vocabulary, where the clothing standard vocabulary includes clothing color, clothing material, clothing type, clothing position, and clothing status.

[0056] Specifically, training the image description model requires language annotation for each image in the training set. In this embodiment of the present invention, to facilitate image comparisons at a higher level of semantics, make it easier for engineers to annotate the training set, and reduce ambiguity, the language annotations do not use formal natural language. Instead, they describe the images using a predefined standard clothing vocabulary.

[0057] In some optional embodiments, the clothing standard vocabulary may include, but is not limited to, parameters such as clothing color, clothing material, clothing type, clothing position, and clothing state that can identify clothing changes in an image. Examples of clothing colors include black, white, gray, red, checkered, striped, and black and red; clothing materials include wool, polyester, silk, denim, and cotton; clothing types include tops, short sleeves, and pants; clothing positions include upper left, upper, upper right, left, center, right, lower left, lower, and lower right; and clothing states include entangled, tumbling, and partially exposed. Compared to commonly used natural languages, these clothing standard vocabulary libraries are smaller in scope and easier to train. Furthermore, these vocabulary libraries are relatively clear, preventing ambiguity among annotators and eliminating the need for multiple people to repeatedly verify and annotate.

[0058] As an optional implementation, the step of obtaining the image description model and the text similarity recognition model can be specifically divided into the following steps S1012 to S1014:

[0059] S1012: Acquire a sample data set, where the sample data set includes a plurality of continuously changing clothing image samples in multiple washing scenes;

[0060] Specifically, the sample data set obtained includes several continuously changing images of clothes in the drum of a washing machine. Different clothing washing scenes can be set for shooting to obtain sample data sets of multiple clothing washing scenes. For example, different materials (such as cotton, chemical fiber, wool, etc.), different colors (light color, dark color, mixed color), different amounts of clothes (small amount, appropriate amount, large amount) and different washing modes (standard wash, fast wash, gentle wash, etc.) are shot. For each washing scene, after the washing machine starts running, continuous shooting is performed at a fixed time interval until the washing program ends. The time interval can be set according to actual needs, such as 1 second, 3 seconds and 4 seconds. During the shooting process, the camera parameters such as focal length, shutter speed, aperture size and ISO are kept unchanged to ensure that the acquired images are consistent in visual features.

[0061] S1013, performing text description on each clothing image sample according to a clothing standard vocabulary to obtain a description text sample corresponding to the clothing image sample;

[0062] It should be noted that in the embodiment of the present invention, the standard clothing vocabulary is restricted to a limited vocabulary, that is, it mainly expresses attribute information such as color, type, material, and the approximate position information of the clothing in the image. Therefore, only a smaller-scale module is needed to construct the image description model, thereby improving the model's computing speed.

[0063] S1014: Input a plurality of clothing image samples into a pre-built first image description model for training, and optimize the parameters of the trained first image description model according to the description text samples to obtain an image description model.

[0064] Specifically, the acquired clothing image samples are input into the constructed first image description model, and the model outputs the descriptive text corresponding to the clothing image samples. Then, the descriptive text samples annotated by engineers are used as a test set. Through the back propagation algorithm, the parameters in the first image description model are updated according to the test set and the descriptive text output by the model to obtain a trained image description model.

[0065] In some optional embodiments, due to computing power limitations, an overly large language model cannot be used. In the laundry recognition scenario in the washing machine drum of an embodiment of the present invention, the focus is on a piece of information about the laundry, that is, some specific vocabulary information. This vocabulary information is used as a simplified "language" to describe the image, and then a very simple vocabulary output model can be constructed to complete the image description task. Figure 2The figure shows a schematic diagram of the structure of the first image description model. The embodiment of the present invention constructs a simple first image description model, which is mainly composed of two parts: a convolutional neural network (CNN) and a long short-term memory network (LSTM). The convolutional neural network (CNN) is used to extract features of clothing images, capture key features and semantic information of clothing images, obtain image features, and generate corresponding description text through the long short-term memory network (LSTM).

[0066] As an optional implementation, the step of obtaining the image description model and the text similarity recognition model can be further divided into the following steps S1015 and S1016:

[0067] S1015, performing similarity comparison on the description text samples of the continuous changes before and after, and obtaining a similarity comparison result;

[0068] Specifically, for the data set preparation of the text similarity recognition model, two clothing image samples that change continuously before and after and the corresponding description text samples are shown to engineering personnel. The engineering personnel give labels, determine the similarity between the two clothing image samples, and obtain the similarity comparison results.

[0069] It should be noted that the presence of descriptive information allows engineers to use a unified standard to determine image similarity. For example, if the clothing in the upper left corner of two images changes, the two images are considered dissimilar. If the two images differ in content but the clothing is roughly located in the same position, the image descriptions are the same, and the two images are considered similar. A slight rotation of the washing machine drum doesn't significantly change the clothing; it simply shifts left or right overall in the image. In this case, the description text generated by the image description model won't change significantly, indicating that the two images are similar based on the description. The presence of image descriptions makes it easier for engineers to determine the degree of image similarity, reducing ambiguity and reducing the workload of repeated labeling.

[0070] S1016: Input a number of description text samples into a pre-built classification model for training, optimize the parameters of the trained classification model according to the similarity comparison results, and obtain a text similarity recognition model.

[0071] Specifically, the structure of the classification model can adopt an LSTM classification model or a classification model based on a transformer structure. The descriptive text samples output by the image description model are input into the constructed classification model, and the model outputs the corresponding classification results. The similarity comparison results compared by the engineering personnel are then used as the test set. Through the back propagation algorithm, the parameters in the classification model are updated according to the test set and the classification results output by the model to obtain a trained text similarity recognition model.

[0072] S102, obtaining a clothing reference image and a clothing image to be identified;

[0073] Specifically, an image of clothes in a washing machine drum is selected as a clothes reference image, ie, an image to be compared, and an image of clothes to be identified is obtained at the same time.

[0074] S103, obtaining a clothing change recognition result based on the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model;

[0075] As a further optional implementation, the step of obtaining a clothing change recognition result based on the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model can be specifically divided into the following steps S1031 to S1033:

[0076] S1031, inputting the clothing reference image into the image description model to obtain a first description text;

[0077] S1032: Input the clothing image to be recognized into the image description model to obtain a second description text;

[0078] S1033: Input the first description text and the second description text into a text similarity recognition model to obtain a clothing change recognition result.

[0079] Specifically, if Figure 3 The figure shows the steps for obtaining clothing change recognition results. A clothing reference image is described using an image description model to obtain a first description text corresponding to the clothing reference image. The clothing image to be recognized is also described using the image description model to obtain a second description text corresponding to the clothing image to be recognized. The first and second description texts obtained are then input into a text similarity recognition model to determine whether the two descriptions are consistent, thereby obtaining the clothing change recognition results. The text similarity recognition model can employ a text classification model or a text grading model.

[0080] As an optional implementation, step S1033 may be specifically divided into the following steps A10331 to A10334:

[0081] A10331. Concatenate the first description text and the second description text in a preset order to obtain a description text vector;

[0082] A10332. Input the description text vector into the text similarity recognition model for classification to obtain similarity categories, including similarity and dissimilarity.

[0083] A10333. When the similarity category is similar, determining that the image of the clothing to be identified has not changed;

[0084] A10334: When the similarity category is dissimilar, determine that the image of the clothing to be identified has changed.

[0085] In some optional embodiments, after obtaining the first description text and the second description text corresponding to the clothing reference image and the clothing image to be identified, the two description texts are spliced ​​into one description, and the order is uniformly stipulated that the description text of the clothing reference image is in front and the description text of the clothing image to be identified is in the back. The spliced ​​description (expressed in the program, that is, a vector) is input into the text similarity recognition model for classification, and the output categories are two categories, namely similarity and dissimilarity, so as to determine whether the clothing image to be identified has changed.

[0086] As an optional implementation, step S1033 may be specifically divided into the following steps B10331 to B10335:

[0087] B10331, set similarity threshold;

[0088] B10332. Concatenate the first description text and the second description text in a preset order to obtain a description text vector;

[0089] B10333. Input the description text vector into the text similarity recognition model for grading to obtain a similarity level;

[0090] B10334. When the similarity level is equal to or greater than the similarity threshold, it is determined that the image of the clothing to be identified has not changed;

[0091] B10335. When the similarity level is less than the similarity threshold, it is determined that the image of the clothing to be identified has changed.

[0092] In some optional embodiments, a similarity threshold is pre-set as a basis for determining whether a clothing image has changed. This similarity threshold can be set based on actual scenario requirements and is not limited here. After obtaining the first and second description texts corresponding to the clothing reference image and the clothing image to be identified, the two description texts are spliced ​​into a description, with the description text of the clothing reference image first and the description text of the clothing image to be identified second. The spliced ​​description (expressed in the program as a vector) is input into a text similarity recognition model for classification, for example, into five levels, ranging from very similar, highly similar, moderately similar, slightly similar, to very dissimilar, to distinguish different levels, thereby determining whether the clothing image to be identified has changed based on the similarity level and the pre-set similarity.

[0093] The above describes the clothing change recognition method according to an embodiment of the present invention. It can be appreciated that, compared to existing image processing methods, the present invention, on the one hand, replaces the direct image similarity comparison method with the method of comparing descriptive text similarity. The clothing reference image and the clothing image to be recognized are described verbally using an image description model to obtain corresponding descriptive text. The obtained descriptive text is then input into a text similarity recognition model to determine whether the contents of the two descriptive texts are consistent, thereby obtaining a clothing change recognition result. This allows for image comparison from a higher-level semantic perspective. Specifically, the description of the image content, derived from semantics, better expresses the main idea of ​​the image, resulting in higher comparison accuracy and a comparison result that is more consistent with human judgment standards. Furthermore, a standard clothing vocabulary is defined for clothing recognition, making the image description model easier to train and smaller in size, thereby ensuring efficient model inference.

[0094] Reference Figure 4 , an embodiment of the present invention further provides a clothing change recognition system, comprising:

[0095] The first module is used to obtain an image description model and a text similarity recognition model;

[0096] The second module is used to obtain clothing reference images and clothing images to be identified;

[0097] The third module is used to obtain clothing change recognition results based on the clothing reference image and the clothing image to be recognized through the image description model and the text similarity recognition model.

[0098] The contents of the above-mentioned clothing change recognition method embodiment are all applicable to the present clothing change recognition system embodiment. The functions specifically implemented by the present clothing change recognition system embodiment are the same as those of the above-mentioned clothing change recognition method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned clothing change recognition method embodiment.

[0099] An embodiment of the present invention further provides an electronic device comprising: a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for enabling communication between the processor and the memory. When the program is executed by the processor, the aforementioned clothing change recognition method is implemented. The electronic device can be any intelligent terminal, including a tablet computer and an in-vehicle computer.

[0100] like Figure 5 FIG2 is a schematic diagram of the hardware structure of the electronic device provided by the embodiment of the present invention, referring to FIG2 Figure 5 , an embodiment of the present invention provides an electronic device, including:

[0101] The processor 1001 may be implemented as a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is configured to execute relevant programs to implement the technical solutions provided by the embodiments of the present invention.

[0102] The memory 1002 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1002 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program codes are stored in the memory 1002 and are called by the processor 1001 to execute the clothing change recognition method of the embodiments of the present invention.

[0103] Input / output interface 1003, used to implement information input and output;

[0104] Communication interface 1004, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0105] Bus 1005 , which transmits information between various components of the device (e.g., processor 1001 , memory 1002 , input / output interface 1003 , and communication interface 1004 );

[0106] The processor 1001 , the memory 1002 , the input / output interface 1003 and the communication interface 1004 are connected to each other in communication within the device via the bus 1005 .

[0107] An embodiment of the present invention also provides a storage medium, which is a computer-readable storage medium used for computer-readable storage. The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above-mentioned clothing change identification method.

[0108] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0109] The embodiment of the present invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs Figure 1 The method shown.

[0110] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the above-mentioned boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operation and logic flow presented herein. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.

[0111] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the above-mentioned functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present invention. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the routine skills of an engineer. Therefore, a person skilled in the art can implement the present invention set forth in the claims using ordinary skills without undue experimentation. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0112] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the above methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0113] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0114] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0115] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

[0116] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for identifying clothing changes, characterized in that: The following steps are involved: Obtain image description model and text similarity recognition model; Acquire a clothing reference image and an image of clothing to be identified; A clothing change recognition result is obtained according to the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model.

2. The method for identifying clothing changes according to claim 1, wherein: The clothing change recognition method further includes: A clothing standard vocabulary is defined, wherein the clothing standard vocabulary includes clothing color, clothing material, clothing type, clothing position, and clothing status.

3. The method for identifying clothing changes according to claim 2, wherein: The acquisition of the image description model and the text similarity recognition model specifically includes: Acquire a sample data set, wherein the sample data set includes a plurality of continuously changing clothing image samples in various washing scenes; Performing a text description on each clothing image sample according to the clothing standard vocabulary to obtain a description text sample corresponding to the clothing image sample; A plurality of clothing image samples are input into a preset first image description model for training, and parameters of the trained first image description model are optimized according to the description text samples to obtain the image description model.

4. The method for identifying clothing changes according to claim 3, wherein: The acquisition of the image description model and the text similarity recognition model specifically includes: Performing similarity comparison on the description text samples that change continuously before and after to obtain a similarity comparison result; A plurality of the description text samples are input into a pre-built classification model for training, and parameters of the trained classification model are optimized according to the similarity comparison result to obtain the text similarity recognition model.

5. The method for identifying clothing changes according to claim 1, wherein: The obtaining of a clothing change recognition result according to the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model specifically includes: Inputting the clothing reference image into the image description model to obtain a first description text; Inputting the clothing image to be identified into the image description model to obtain a second description text; The first description text and the second description text are input into the text similarity recognition model to obtain the clothing change recognition result.

6. The method for identifying clothing changes according to claim 5, characterized in that: Inputting the first description text and the second description text into the text similarity recognition model to obtain the clothing change recognition result specifically includes: Concatenate the first description text and the second description text in a preset order to obtain a description text vector; Inputting the description text vector into the text similarity recognition model for classification to obtain similarity categories, wherein the similarity categories include similarity and dissimilarity; When the similarity category is similar, determining that the image of the clothing to be identified has not changed; When the similarity category is dissimilar, it is determined that the to-be-identified clothing image has changed.

7. The method for identifying clothing changes according to claim 5, wherein: Inputting the first description text and the second description text into the text similarity recognition model to obtain the clothing change recognition result specifically includes: Set similarity threshold; Concatenate the first description text and the second description text in a preset order to obtain a description text vector; Inputting the description text vector into the text similarity recognition model for grading to obtain a similarity level; When the similarity level is equal to or greater than the similarity threshold, determining that the image of the clothing to be identified has not changed; When the similarity level is less than the similarity threshold, it is determined that the to-be-recognized clothing image has changed.

8. A clothing change recognition system, characterized in that: include: The first module is used to obtain an image description model and a text similarity recognition model; The second module is used to obtain clothing reference images and clothing images to be identified; The third module is configured to obtain a clothing change recognition result based on the clothing reference image and the clothing image to be recognized by using the image description model and the text similarity recognition model.

9. An electronic device, characterized in that: The electronic device includes a memory, a processor, a program stored in the memory and executable on the processor, and a data bus for realizing connection and communication between the processor and the memory. When the program is executed by the processor, the steps of the clothing change identification method described in any one of claims 1 to 7 are realized.

10. A storage medium, which is a computer-readable storage medium and is used for computer-readable storage, characterized in that: The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the clothing change recognition method according to any one of claims 1 to 7.

Images