CN106295566A - Facial expression recognizing method and device - Google Patents

Abstract

The present disclosure relates to a method and device for recognizing facial expressions, and belongs to the technical field of image recognition. The method includes: detecting a face area from an image to be recognized; obtaining key points in the face area; extracting a partial image from the face area according to the key points; using a trained expression recognition model to identify the partial image, and obtaining Expression recognition results. The present disclosure combines the two separate steps into one by using the trained facial expression recognition model for feature extraction and facial expression discrimination, thereby reducing cumulative errors and improving the accuracy of facial facial expression recognition. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

Description

Facial expression recognition method and device

technical field

The present disclosure relates to the technical field of image recognition, in particular to a method and device for recognizing facial expressions.

Background technique

Facial expression recognition refers to identifying and determining the expression state of a human face from a given facial image. For example, happy, sad, surprised, fearful, disgusted, angry, etc. At present, facial expression recognition has been widely used in psychological science, neuroscience, engineering science, computer science and other fields.

In the related art, facial expression recognition includes the following two main steps: first, detect the human face area from the image to be recognized, and extract the facial expression features from the human face area, wherein, HOG (Histogram of OrientedGradient , directional gradient histogram), LBP (Local Binary Pattern, local binary pattern), Gabor and other feature extraction algorithms to extract facial expression features; second, perform expression classification based on facial expression features, and obtain expression recognition results, wherein, classification The algorithm can adopt Adaboost algorithm, SVM (Support Vector Machine, support vector machine) algorithm, random forest algorithm, etc.

Contents of the invention

Embodiments of the present disclosure provide a method and device for recognizing facial expressions. Described technical scheme is as follows:

According to a first aspect of an embodiment of the present disclosure, a method for recognizing facial expressions is provided, the method comprising:

Detect the face area from the image to be recognized;

Obtain key points in the face area;

Extracting a partial image from the face area according to the key point;

The partial image is recognized by using the trained expression recognition model to obtain an expression recognition result.

Optionally, the extracting a partial image from the face region according to the key points includes:

Get the image blocks around each key point;

The partial image is obtained by superimposing or splicing each of the image blocks in a preset order.

Optionally, the acquiring image blocks around each key point includes:

For each key point, an image block of a predetermined size centered on the key point is intercepted.

Optionally, using the trained expression recognition model to identify the partial image to obtain an expression recognition result, including:

Extracting feature information of the partial image by using the facial expression recognition model that has been trained;

Using the expression recognition model to determine the expression recognition result according to the feature information.

Optionally, the acquiring key points in the face area includes:

Scaling the face area to a target size;

Locating the key points from the zoomed face area.

Optionally, the facial expression recognition model is a convolutional neural network model.

According to a second aspect of an embodiment of the present disclosure, a human facial expression recognition device is provided, the device comprising:

A face detection module configured to detect a face area from an image to be recognized;

a key point acquisition module configured to acquire key points in the face area;

An image extraction module configured to extract a partial image from the face area according to the key points;

The expression recognition module is configured to use the trained expression recognition model to recognize the partial image to obtain an expression recognition result.

Optionally, the image extraction module includes:

The image block acquisition sub-module is configured to acquire image blocks around each key point;

The image block processing sub-module is configured to superimpose or stitch each of the image blocks in a preset order to obtain the partial image.

Optionally, the image block acquisition submodule is configured to, for each key point, intercept an image block of a predetermined size centered on the key point.

Optionally, the facial expression recognition module includes:

The feature extraction submodule is configured to extract feature information of the partial image using the trained expression recognition model;

The identification determination submodule is configured to determine the expression recognition result by using the expression recognition model according to the feature information.

Optionally, the key point acquisition module includes:

The face scaling submodule is configured to scale the face area to a target size;

The key point positioning submodule is configured to locate the key points from the zoomed face area.

Optionally, the facial expression recognition model is a convolutional neural network model.

According to a third aspect of an embodiment of the present disclosure, a human facial expression recognition device is provided, the device comprising:

processor;

memory for storing executable instructions of the processor;

Wherein, the processor is configured as:

Detect the face area from the image to be recognized;

Obtain key points in the face area;

Extracting a partial image from the face area according to the key point;

The partial image is recognized by using the trained expression recognition model to obtain an expression recognition result.

The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

By detecting the face area from the image to be recognized, the key points in the face area are obtained, and the partial image is extracted from the face area according to the key point, and the partial image is recognized by the trained expression recognition model, and the expression recognition is obtained. Results: Solve the problem that in related technologies, since feature extraction and expression discrimination are two separate steps, two different algorithms need to be used for processing, resulting in cumulative errors and affecting the accuracy of facial expression recognition; The expression recognition model performs feature extraction and expression discrimination at key points of the face, and combines the two separate steps into one, thereby reducing cumulative errors and improving the accuracy of facial expression recognition. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description serve to explain the principles of the disclosure.

Fig. 1 is a flow chart of a method for recognizing facial expressions according to an exemplary embodiment;

Fig. 2A is a flowchart of a method for recognizing facial expressions according to another exemplary embodiment;

Fig. 2B is a schematic diagram of key point positioning involved in the embodiment shown in Fig. 2A;

Fig. 2C is a schematic structural diagram of a convolutional neural network shown exemplarily;

Fig. 3 is a block diagram of a human facial expression recognition device shown according to an exemplary embodiment;

Fig. 4 is a block diagram of a human facial expression recognition device according to another exemplary embodiment;

Fig. 5 is a block diagram of a device according to an exemplary embodiment.

detailed description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatuses and methods consistent with aspects of the present disclosure as recited in the appended claims.

In related technologies, feature extraction and expression discrimination are two separate steps, which need to be processed by two different algorithms respectively, and there are accumulated errors, which affect the accuracy of facial expression recognition. In the technical solution provided by the embodiments of the present disclosure, the trained expression recognition model is used to perform feature extraction and expression discrimination at key points of the face, and the two separate steps are combined into one, thereby reducing cumulative errors and improving facial expression recognition accuracy. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

In the method provided by the embodiments of the present disclosure, each step may be executed by an electronic device capable of image processing, such as a personal computer, a smart phone, a tablet computer, a server, and the like. For the convenience of description, in the following method embodiments, the execution subject of each step is an electronic device for description.

Fig. 1 is a flow chart of a method for recognizing facial expressions according to an exemplary embodiment. The method may include the following steps:

In step 101, a human face area is detected from an image to be recognized.

In step 102, key points in the face area are acquired.

Key points, also known as feature points, face key points, or face feature points, refer to the facial positions in the face area that can reflect the state of expression, including but not limited to eyes (such as eye corners, eyeball centers, and eye ends), noses, etc. (such as nose tip, nose wing), mouth (such as mouth corner, lip corner, lip edge), chin, eyebrow corner and other facial positions.

In step 103, a partial image is extracted from the face area according to the key points.

In step 104, the trained facial expression recognition model is used to recognize the partial image to obtain the facial expression recognition result.

To sum up, the method provided in this embodiment solves the problem that in related technologies, since feature extraction and expression discrimination are two separate steps, two different algorithms need to be used for processing, resulting in cumulative errors that affect facial expressions. The problem of recognition accuracy. The trained expression recognition model is used for feature extraction and expression discrimination at key points of the face, and the two separate steps are combined into one, thereby reducing cumulative errors and improving the accuracy of facial expression recognition. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

Fig. 2A is a flow chart of a method for recognizing facial expressions according to another exemplary embodiment. The method may include the following steps:

In step 201, a human face area is detected from an image to be recognized.

The electronic device uses a related face detection algorithm to detect the face area from the image to be recognized. In this embodiment, the specific type of the face detection algorithm is not limited. For example, the face detection algorithm can be a combination of the LBP algorithm and the Adaboost algorithm. The LBP algorithm is used to extract image features from the image to be recognized, and the Adaboost cascade classifier is used to determine the face area according to the image features. The image to be recognized may be an image containing a human face.

In step 202, the face area is scaled to a target size.

Since the size of the face area in different images is different, in order to ensure the accuracy of the subsequent feature point positioning, the extracted face area is scaled to a fixed target size. In this embodiment, there is no limitation on the size of the target size, which can be preset according to the actual situation. For example, the target size is 96×96 (pixels).

In step 203, key points are located from the zoomed face area.

Key points, also known as feature points, face key points, or face feature points, refer to the facial positions in the face area that can reflect the state of expression, including but not limited to eyes (such as eye corners, eyeball centers, and eye ends), noses, etc. (such as nose tip, nose wing), mouth (such as mouth corner, lip corner, lip edge), chin, eyebrow corner and other facial positions.

The electronic device locates the key points from the zoomed face area by using a relevant human face key point positioning algorithm. In this embodiment, there is no limitation on the specific type of facial key point location algorithm. For example, the facial key point location algorithm may be an SDM (Supervised Descent Method, supervised descent method) algorithm.

In one example, as shown in FIG. 2B , the SDM algorithm is used to locate and obtain multiple key points from the zoomed face area 21 , and each key point is shown as a small black dot in the figure. The position and quantity of the key points to be located can be preset, for example, 20.

In step 204, image blocks around each key point are obtained.

The image blocks around the key point refer to image blocks of a predetermined size including the key point. In one example, for each key point, an image block of a predetermined size centered on the key point is intercepted. In this embodiment, there is no limitation on the size of the predetermined size, which can be preset according to the actual situation. For example, the predetermined size is 32×32 (pixels).

Optionally, before performing key point positioning, the face area can be converted into a grayscale image, and key points can be located in the grayscale image. Correspondingly, the image blocks around the intercepted key points are also grayscale images.

In step 205, each image block is superimposed or spliced in a preset order to obtain a partial image.

Before facial expression recognition, each image block needs to be integrated and input to the facial expression recognition model as a whole. In a possible implementation manner, each image block is superimposed in a preset order and used as an input to the expression recognition model. In another possible implementation manner, each image block is spliced in a preset order and used as an input to the expression recognition model.

Since there are usually multiple key points acquired, and different key points correspond to different facial positions, each image block needs to be superimposed or spliced in a preset order. Taking key points including eyes, nose, mouth, eyebrows, and chin as an example, the preset order can be eyes, nose, mouth, chin, and eyebrows. That is to say, whether it is in the training stage of the expression recognition model or in the expression recognition stage using the expression recognition model, multiple image blocks intercepted from any image are superimposed or spliced according to the same preset order, so that Ensure that the structure of the input data of the facial expression recognition model is unified.

In step 206, the trained facial expression recognition model is used to recognize the partial image to obtain the facial expression recognition result.

The electronic device uses the trained facial expression recognition model to extract feature information of the partial image, and uses the facial expression recognition model to determine the facial expression recognition result based on the feature information. In this embodiment, feature extraction and expression discrimination are both completed by the expression recognition model, and there is no need to use two different algorithms to complete it in two steps. The two separate steps are combined into one, thereby reducing cumulative errors and improving facial recognition. Accuracy of expression recognition.

In one example, the facial expression recognition model is a convolutional neural network (Convolutional Neural Network, CNN) model, or called a deep convolutional neural network model. The convolutional neural network has powerful feature extraction capabilities, and the use of convolutional neural networks for feature extraction and expression discrimination has high accuracy. A convolutional neural network includes an input layer, at least one convolutional layer, at least one fully connected layer, and an output layer. Among them, the input data of the input layer is the local image obtained by superimposing or splicing each image block in sequence; the output result of the output layer is a vector of length n, respectively representing the probability of n kinds of expressions, and n is an integer greater than 1 . Convolutional layers are used for feature extraction. The fully connected layer is used to combine and abstract the features extracted by the convolutional layer to obtain data suitable for classification by the output layer.

Referring to FIG. 2C , it exemplarily shows a schematic structural diagram of a convolutional neural network. The convolutional neural network includes 1 input layer, 3 convolutional layers (the first convolutional layer C1, the second convolutional layer C2 and the third convolutional layer C3), 2 fully connected layers (the first fully connected layer FC4 and the second fully connected layer FC5) and 1 output layer (Softmax layer). Assuming that 20 key points are extracted from the face area, the size of the image block around each key point is 32×32 (pixels), and the input data of the input layer is a superimposed image or splicing of 20×32×32 image. The numbers of convolution kernels in the three convolutional layers C1, C2 and C3 are 36, 64 and 32 respectively. The step size of the first convolutional layer C1 is 2. After the calculation of the first convolutional layer C1, the length and width of the image are reduced to half of the original. The second convolutional layer C2 and the third convolutional layer C3 The step size is 1. It should be noted that the convolutional neural network shown in FIG. 2C is only exemplary and explanatory, and is not intended to limit the present disclosure. Generally speaking, the more layers of the convolutional neural network, the better the effect but the longer the calculation time. In practical applications, a convolutional neural network with an appropriate number of layers can be designed in combination with the requirements for recognition accuracy and efficiency.

In summary, the method provided in this embodiment uses the trained facial expression recognition model to perform feature extraction and facial expression discrimination at key points of the face, and combines the two separate steps into one, thereby reducing cumulative errors and improving human facial expressions. Accuracy of facial expression recognition. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

The following are device embodiments of the present disclosure, which can be used to implement the method embodiments of the present disclosure. For details not disclosed in the disclosed device embodiments, please refer to the disclosed method embodiments.

Fig. 3 is a block diagram of a device for recognizing facial expressions according to an exemplary embodiment. The device has the function of realizing the above method example, and the function may be realized by hardware, or may be realized by executing corresponding software by hardware. The device may include: a face detection module 310 , a key point acquisition module 320 , an image extraction module 330 and an expression recognition module 340 .

The face detection module 310 is configured to detect a face area from the image to be recognized.

The key point acquiring module 320 is configured to acquire the key points in the face area.

The image extraction module 330 is configured to extract a partial image from the face area according to the key points.

The expression recognition module 340 is configured to use the trained expression recognition model to recognize the partial image to obtain an expression recognition result.

In summary, the device provided in this embodiment uses the trained facial expression recognition model to perform feature extraction and facial expression discrimination at key points of the face, and combines the two separate steps into one, thereby reducing cumulative errors and improving human facial expression. Accuracy of facial expression recognition. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

Fig. 4 is a block diagram of a device for recognizing facial expressions according to another exemplary embodiment. The device has the function of realizing the above method example, and the function may be realized by hardware, or may be realized by executing corresponding software by hardware. The device may include: a face detection module 310 , a key point acquisition module 320 , an image extraction module 330 and an expression recognition module 340 .

The face detection module 310 is configured to detect a face area from the image to be recognized.

The key point acquiring module 320 is configured to acquire the key points in the face area.

The image extraction module 330 is configured to extract a partial image from the face area according to the key points.

The expression recognition module 340 is configured to use the trained expression recognition model to recognize the partial image to obtain an expression recognition result.

In an example, the image extraction module 330 includes: an image block acquisition submodule 330a and an image block processing submodule 330b.

The image block acquisition sub-module 330a is configured to acquire image blocks around each key point.

The image block processing sub-module 330b is configured to superimpose or stitch each of the image blocks in a preset order to obtain the partial image.

In one example, the image block acquisition sub-module 330b is configured to, for each key point, intercept an image block of a predetermined size centered on the key point.

In one example, the expression recognition module 340 includes: a feature extraction submodule 340a and a recognition determination submodule 340b.

The feature extraction sub-module 340a is configured to use the trained expression recognition model to extract feature information of the partial image.

The recognition determination sub-module 340b is configured to determine the expression recognition result by using the expression recognition model according to the feature information.

In an example, the key point acquisition module 320 includes: a face scaling submodule 320a and a key point positioning submodule 320b.

The human face scaling sub-module 320a is configured to scale the human face area to a target size.

The key point positioning sub-module 320b is configured to locate the key points from the zoomed face area.

In one example, the facial expression recognition model is a convolutional neural network model.

In summary, the device provided in this embodiment uses the trained facial expression recognition model to perform feature extraction and facial expression discrimination at key points of the face, and combines the two separate steps into one, thereby reducing cumulative errors and improving human facial expression. Accuracy of facial expression recognition. Moreover, because only the feature information of the local image at the key points of the face is extracted, rather than the global feature information of the entire face area, the features that reflect the state of expression can be extracted more accurately and effectively, and the performance of facial expression recognition can be further improved. accuracy.

It should be noted that when the device provided by the above embodiment realizes its functions, it only uses the division of the above-mentioned functional modules as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional modules according to actual needs. That is, the content structure of the device is divided into different functional modules to complete all or part of the functions described above.

Regarding the apparatus in the foregoing embodiments, the specific manner in which each module executes operations has been described in detail in the embodiments related to the method, and will not be described in detail here.

An exemplary embodiment of the present disclosure also provides a human facial expression recognition device capable of implementing the human facial expression recognition method provided in the present disclosure. The device includes: a processor, and a memory for storing executable instructions of the processor. where the processor is configured as:

Detect the face area from the image to be recognized;

Obtain key points in the face area;

Extracting a partial image from the face area according to the key point;

The partial image is recognized by using the trained expression recognition model to obtain an expression recognition result.

Optionally, the processor is configured to:

Get image blocks around each key point;

The partial image is obtained by superimposing or splicing each of the image blocks in a preset order.

Optionally, the processor is configured to:

For each key point, an image block of a predetermined size centered on the key point is intercepted.

Optionally, the processor is configured to:

Extracting feature information of the partial image by using the facial expression recognition model that has been trained;

Using the expression recognition model to determine the expression recognition result according to the feature information.

Optionally, the processor is configured to:

Scaling the face area to a target size;

Locating the key points from the zoomed face area.

Optionally, the facial expression recognition model is a convolutional neural network model.

Fig. 5 is a block diagram of an apparatus 500 according to an exemplary embodiment. For example, the apparatus 500 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, and the like.

5, apparatus 500 may include one or more of the following components: processing component 502, memory 504, power supply component 506, multimedia component 508, audio component 510, input/output (I/O) interface 512, sensor component 514, and Communication component 516 .

The processing component 502 generally controls the overall operations of the device 500, such as those associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 502 may include one or more processors 520 to execute instructions to complete all or part of the steps of the above method. Additionally, processing component 502 may include one or more modules that facilitate interaction between processing component 502 and other components. For example, processing component 502 may include a multimedia module to facilitate interaction between multimedia component 508 and processing component 502 .

The memory 504 is configured to store various types of data to support operations at the device 500 . Examples of such data include instructions for any application or method operating on device 500, contact data, phonebook data, messages, pictures, videos, and the like. The memory 504 can be implemented by any type of volatile or non-volatile storage device or their combination, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.

The power supply component 506 provides power to various components of the device 500 . Power components 506 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for device 500 .

The multimedia component 508 includes a screen that provides an output interface between the device 500 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may not only sense a boundary of a touch or swipe action, but also detect duration and pressure associated with the touch or swipe action. In some embodiments, the multimedia component 508 includes a front camera and/or a rear camera. When the device 500 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capability.

The audio component 510 is configured to output and/or input audio signals. For example, the audio component 510 includes a microphone (MIC), which is configured to receive external audio signals when the device 500 is in operation modes, such as call mode, recording mode and voice recognition mode. Received audio signals may be further stored in memory 504 or sent via communication component 516 . In some embodiments, the audio component 510 also includes a speaker for outputting audio signals.

The I/O interface 512 provides an interface between the processing component 502 and a peripheral interface module. The peripheral interface module may be a keyboard, a click wheel, a button, and the like. These buttons may include, but are not limited to: a home button, volume buttons, start button, and lock button.

Sensor assembly 514 includes one or more sensors for providing status assessments of various aspects of device 500 . For example, the sensor component 514 can detect the open/closed state of the device 500, the relative positioning of components, such as the display and keypad of the device 500, and the sensor component 514 can also detect a change in the position of the device 500 or a component of the device 500 , the presence or absence of user contact with the device 500 , the device 500 orientation or acceleration/deceleration and the temperature change of the device 500 . Sensor assembly 514 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact. Sensor assembly 514 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor component 514 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor or a temperature sensor.

The communication component 516 is configured to facilitate wired or wireless communication between the apparatus 500 and other devices. The device 500 can access wireless networks based on communication standards, such as Wi-Fi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 516 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 516 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wide Band (UWB) technology, Bluetooth (BT) technology and other technologies.

In an exemplary embodiment, apparatus 500 may be programmed by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation for performing the methods described above.

In an exemplary embodiment, there is also provided a non-transitory computer-readable storage medium including instructions, such as the memory 504 including instructions, which can be executed by the processor 520 of the device 500 to implement the above method. For example, the non-transitory computer readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like.

A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by the processor of the device 500, the device 500 can execute the above method.

It should be understood that the "plurality" mentioned herein refers to two or more than two. "And/or" describes the association relationship of associated objects, indicating that there may be three types of relationships, for example, A and/or B may indicate: A exists alone, A and B exist simultaneously, and B exists independently. The character "/" generally indicates that the contextual objects are an "or" relationship.

Other embodiments of the present disclosure will be readily apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present disclosure, and these modifications, uses or adaptations follow the general principles of the present disclosure and include common knowledge or conventional technical means in the technical field not disclosed in the present disclosure . The specification and examples are to be considered exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It should be understood that the present disclosure is not limited to the precise constructions which have been described above and shown in the drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims (13)

1. a facial expression recognition method, is characterized in that, described method comprises: Detect the face area from the image to be recognized; Obtain key points in the face area; Extracting a partial image from the face area according to the key point; The partial image is recognized by using the trained expression recognition model to obtain an expression recognition result. 2. The method according to claim 1, wherein said extracting a partial image from said face region according to said key points comprises: Get image blocks around each key point; The partial image is obtained by superimposing or splicing each of the image blocks in a preset order. 3. The method according to claim 2, wherein said acquiring image blocks around each key point comprises: For each key point, an image block of a predetermined size centered on the key point is intercepted. 4. method according to claim 1, is characterized in that, described partial image is identified by the facial expression recognition model that described adopting finishing training, obtains facial expression recognition result, comprises: Extracting feature information of the partial image by using the facial expression recognition model that has been trained; Using the expression recognition model to determine the expression recognition result according to the characteristic information. 5. method according to claim 1, is characterized in that, described obtaining the key point in described people's face area, comprises: Scaling the face area to a target size; Locating the key points from the zoomed face area. 6. The method according to any one of claims 1 to 5, wherein the facial expression recognition model is a convolutional neural network model. 7. A facial expression recognition device, characterized in that said device comprises: A face detection module configured to detect a face area from an image to be recognized; a key point acquisition module configured to acquire key points in the face area; An image extraction module configured to extract a partial image from the face area according to the key points; The expression recognition module is configured to use the trained expression recognition model to recognize the partial image to obtain an expression recognition result. 8. The device according to claim 7, wherein the image extraction module comprises: The image block acquisition sub-module is configured to acquire image blocks around each key point; The image block processing sub-module is configured to superimpose or stitch each of the image blocks in a preset order to obtain the partial image. 9. The device of claim 8, wherein: The image block acquisition sub-module is configured to, for each key point, intercept an image block of a predetermined size centered on the key point. 10. The device according to claim 7, wherein the facial expression recognition module comprises: The feature extraction submodule is configured to extract feature information of the partial image using the trained expression recognition model; The identification determination submodule is configured to determine the expression recognition result by using the expression recognition model according to the feature information. 11. The device according to claim 7, wherein the key point acquisition module comprises: The face scaling submodule is configured to scale the face area to a target size; The key point positioning submodule is configured to locate the key points from the zoomed face area. 12. The device according to any one of claims 7 to 11, wherein the facial expression recognition model is a convolutional neural network model. 13. A facial expression recognition device, characterized in that the device comprises: processor; memory for storing executable instructions of the processor; Wherein, the processor is configured as: Detect the face area from the image to be recognized; Obtain key points in the face area; Extracting a partial image from the face area according to the key point; The partial image is recognized by using the trained expression recognition model to obtain an expression recognition result.