Calorie estimation method, device, equipment and storage medium
A deep learning-based method estimates calorie consumption in training courses by analyzing user physical signs and video frames, addressing the inability of existing methods to break down calorie consumption into individual exercises without heart rate monitoring, enabling precise calorie estimation for smart training plans.
Patent Information
- Application Number
- CN202011589806.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-12-29
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2040-12-29
AI Technical Summary
Existing calorie estimation methods for training courses cannot accurately break down calorie consumption into individual exercises without heart rate monitoring devices, hindering the creation of smart training plans and self-compiled courses.
A calorie estimation method using a deep learning network model that analyzes user physical sign information and training course video frames to determine action completion rates and calorie consumption, employing a standard training action library and physical sign information to estimate calories based on user-specific parameters.
Enables accurate calorie estimation for training courses without heart rate monitoring, allowing for precise calculation of calorie consumption in each exercise and overall training course.
Smart Images

Figure CN112686138B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to the field of computer technology, and in particular to a calorie estimation method, device, equipment and storage medium. Background Art
[0002] As people pay more attention to their health, healthy diet and training courses have become the basic ways to maintain good health. Nutritional elements are taken in through diet and calorie intake is controlled; calories are consumed through training courses to improve body functions. Scientific training courses require accurate estimation of the calories consumed by each training action in the training course. Generally, the calorie calculation of training courses is based on heart rate, training time and various physical indicators of the trainee to estimate calories.
[0003] Since users do not necessarily wear heart rate monitoring devices during training, the calories of exercise training cannot be estimated without heart rate monitoring devices. The existing technology estimates calories by calculating the calories of the entire course, but using the calorie calculation of the entire course to estimate calories cannot be broken down into each exercise in the training course, resulting in the inability to estimate calories for smart training plans and self-compiled courses. Summary of the invention
[0004] The present invention provides a calorie estimation method, device, equipment and storage medium to achieve accurate estimation of the calorie value corresponding to a user's training course based on the user's physical sign information and a training course video frame sequence.
[0005] In a first aspect, an embodiment of the present invention provides a calorie estimation method, the method comprising:
[0006] According to the user's physical sign information, the user's training course video frame sequence and the standard training action library, the action completion rate of each frame image in the training course video frame sequence is obtained;
[0007] The user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through the sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images;
[0008] According to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the total consumed calories of the training course video frame sequence are obtained.
[0009] Furthermore, obtaining the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library includes:
[0010] Determine the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information;
[0011] Determine the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library;
[0012] The completion rate of the user's completed action in each frame in the training course video frame sequence is determined according to the completion parameters of the user's completed matching standard training action and the completion parameters of the action in each frame in the training course video frame sequence.
[0013] Furthermore, the determining of the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information includes:
[0014] Based on the height, weight, limb proportions, and active joints in the trainee's physical information in the standard training action library and the height, weight, limb proportions, and active joints in the user's physical information, the proportion parameters corresponding to the training actions in the standard training action library completed by the user are determined.
[0015] Furthermore, the determining of the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library includes:
[0016] Determining, according to the training course video frame sequence, the action features of each frame image in the training course video frame sequence and the completion parameters of the user action;
[0017] If a corresponding standard training action is matched with the action features of a standard training action in the standard training action library according to the action features of each frame image in the training course video frame sequence, then the completion parameters of the user completing the matched standard training action are determined according to the completion parameters of the corresponding standard training action, the proportion parameters corresponding to the training action in the standard training action library completed by the user, and the user's vital sign information.
[0018] Furthermore, the calorie estimation model is obtained by training a deep learning network model using sample user vital signs information in a sample training course video frame sequence, two adjacent frames, action completion rates corresponding to the two adjacent frames, and calories consumed by a training action on a previous frame in two adjacent frames. The model includes:
[0019] Acquire a sample training course video frame sequence, training data corresponding to the sample training course video frame sequence, and sample user vital sign information in the sample training course video frame sequence;
[0020] Determine the action completion rate of each frame image in the sample training course video frame sequence according to the action features of each frame image in the sample training course video frame sequence, the action features of the training actions in the standard training action library, the sample user's physical sign information in the sample training course video frame sequence and the trainer feature information in the standard training library;
[0021] Determine, based on the training data corresponding to the sample training course video frame sequence and the sample training course video frame sequence, the calories consumed by changing from a training action on a previous frame image to a training action on a subsequent frame image in two adjacent frames of the sample training course video frame sequence;
[0022] The two adjacent frames of images in the sample training course video frame sequence, the sample user's vital signs information in the sample training course video frame sequence, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame image changed to the training action on the next frame image in the two adjacent frames of images in the sample training course video frame sequence are input into the deep learning network model to obtain a calorie estimation model.
[0023] Furthermore, before obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the method further includes:
[0024] If the action feature of the first frame image in the training course video frame sequence matches the training start action in the standard training action library, the calories corresponding to the training start action in the standard training library are counted into the total consumed calories of the training course video frame sequence.
[0025] Furthermore, before obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the method further includes:
[0026] If the action features of the first frame image in the training course video frame sequence match the training start action in the standard training action library, the calories corresponding to the matching action in the standard training action library based on the action features of the first frame image in the training course video frame sequence will be counted into the total consumed calories of the training course video frame sequence.
[0027] In a second aspect, an embodiment of the present invention further provides a calorie estimation device, the device comprising:
[0028] A completion rate determination module, used to obtain the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library;
[0029] A calorie determination module is used to input the user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images;
[0030] The total consumption confirmation module is used to obtain the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence.
[0031] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:
[0032] one or more processors;
[0033] a storage device for storing one or more programs,
[0034] When the one or more programs are executed by the one or more processors, the one or more processors implement any of the above-described calorie estimation methods.
[0035] In a fourth aspect, an embodiment of the present invention further provides a storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform any of the calorie estimation methods when executed by a computer processor.
[0036] The present invention obtains the action completion rate of each frame image in the training course video frame sequence according to user vital sign information, the user's training course video frame sequence and a standard training action library; the user vital sign information, two adjacent frames in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame image changing to the next frame image in the two adjacent frames; the calorie estimation model is obtained by training a deep learning network model through sample user vital sign information, two adjacent frames, the action completion rate corresponding to the two adjacent frames, and the calories consumed by the training action on the previous frame image changing to the next frame image in the two adjacent frames in the sample training course video frame sequence; and the total consumed calories of the training course video frame sequence are obtained according to the calories consumed by the training action on the previous frame image changing to the next frame image in the training course video frame sequence except the first frame image. This solution solves the problem that smart training plans and self-edited courses cannot estimate calories when there is no heart rate collection device, and achieves accurate estimation of the calorie value corresponding to the user's training course based on the user's physical sign information and the training course video frame sequence. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 is a flow chart of a calorie estimation method in Embodiment 1 of the present invention;
[0038] Figure 2 is a flow chart of a calorie estimation method in Embodiment 2 of the present invention;
[0039] Figure 3 is a schematic diagram of the structure of a calorie estimation device in Embodiment 3 of the present invention;
[0040] Figure 4 It is a structural schematic diagram of an electronic device in Embodiment 4 of the present invention. DETAILED DESCRIPTION
[0041] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.
[0042] Embodiment 1
[0043] Figure 1This is a flowchart of a calorie estimation method provided in Embodiment 1 of the present invention. The method can be applied to various scenarios where calorie estimation is required, based on a training course video frame sequence. The method is executed by a calorie estimation device, which can be implemented by software and / or hardware methods, and can be specifically inherited in an electronic device with storage and computing capabilities for calorie estimation.
[0044] like Figure 1 As shown, a calorie estimation method is provided, comprising:
[0045] Step S110, obtaining the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library;
[0046] In an embodiment of the present invention, the user's physical sign information can be understood as the training user's height, age, weight, gender, limb length, limb proportion, active joints and other related information in the training course video frame sequence. The user's training course video frame sequence can be understood as the video corresponding to the training course shot during the user's training process is split into image frames one by one, and these image frames are arranged in the order of shooting time. The standard training action library can be understood as a database added according to the standard action training course video frame sequence in the big database. Among them, the standard training action library includes: each frame image in the standard training course video frame sequence in the big database, the action features of each frame image in the standard training course video frame sequence in the big database, and the trainer's physical sign information. The action completion rate of each frame image in the training course video frame sequence can be understood as the completion percentage of the action of each frame image in the training course video frame sequence to the matching action in the standard training action library calculated based on the user's physical sign information and the trainer's physical sign information.
[0047] In the embodiment of the present invention, the action of each frame image in the user's training course video frame sequence is determined according to the user's physical sign information and the user's training course video frame sequence. The completion parameters corresponding to the training actions in the standard training action library completed by the user are determined according to the user's physical sign information and the trainer information in the standard training action library; the action completion rate of each frame image in the user's training course video frame sequence is determined according to the action completion parameters of each frame image in the user's training course video frame sequence and the completion parameters corresponding to the training actions in the standard training action library completed by the user.
[0048] Step 120, inputting the user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through the sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images;
[0049] In an embodiment of the present invention, two adjacent frames in a training course video frame sequence can be understood as two adjacent frames arranged in the order of shooting time among a plurality of frames in the training course video frame sequence. The action completion rate corresponding to two adjacent frames can be understood as the completion rate corresponding to two adjacent frames arranged in the order of shooting time among a plurality of images in the training course video frame sequence. The calorie estimation model can be understood as an estimation model for calculating the calories consumed by the training action on the previous frame in two adjacent frames according to the user's physical sign information, two adjacent frames in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames. The training action on the previous frame can be understood as an image frame in the training course video frame sequence whose shooting time is earlier than that of the latter frame. The training action on the latter frame can be understood as an image frame in the training course video frame sequence whose shooting time is later than that of the previous frame. The calories consumed by the training action on the previous frame image and the training action on the next frame image in two adjacent frames can be understood as the calorie value consumed by the user when the training action on the previous frame image is changed to the training action on the next frame image.
[0050] In an embodiment of the present invention, the user's physical sign information, two adjacent frames of images in a training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images. If there are a total of N frames of images in the training course video frame sequence, there are N-1 pairs of adjacent frames of images in the training course video frame sequence, and it is necessary to input N-1 adjacent frames of images to calculate the calories consumed by the training action on the previous frame of the two adjacent frames of images corresponding to the adjacent frames of the training course video frame sequence.
[0051] Step S130, obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence changing to the training action on the next frame image.
[0052] In the embodiment of the present invention, the first frame image can be understood as the first frame image in the training course video frame sequence. The total calories consumed in the training course video frame sequence can be understood as the total calories consumed from the training action of the user on the first frame image frame to the training action on the last frame image.
[0053] In the embodiment of the present invention, if two adjacent frames of images in the training course video frame sequence correspond to N frames of images, it is necessary to sequentially input two adjacent frames of images N-1 times to calculate the calories consumed by the training action on the previous frame of the adjacent frames corresponding to the two adjacent frames of images in the N-1 training course video frame sequences. The calculated calories consumed by the training action on the previous frame of the adjacent frames corresponding to the two adjacent frames of images in the N-1 training course video frame sequences are accumulated to obtain the total calories consumed by the training course video frame sequence.
[0054] Furthermore, obtaining the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library includes:
[0055] Determine the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information;
[0056] Determine the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library;
[0057] The completion rate of the user's completed action in each frame in the training course video frame sequence is determined according to the completion parameters of the user's completed matching standard training action and the completion parameters of the action in each frame in the training course video frame sequence.
[0058] In the embodiment of the present invention, the proportional parameters corresponding to the training actions in the standard training action library completed by the user can be understood as the proportional parameters such as the spatial angle formed by the user's limbs, trunk, active joints, horizontal ground, user's center of gravity position and head at each active joint, the distance from the horizontal plane and the user's center of gravity position when the user completes the training action in the standard training action library. The action characteristics of the standard training actions in the standard training action library can be understood as the action attribute information of the standard training actions in the standard training action library, for example: the degree of deviation of the user's center of gravity position, the involved limbs in the training action, the spatial angle formed by the active joints and the distance between the limbs and the horizontal plane. The action characteristics of each frame image in the training course video frame sequence can be understood as the action attribute information of the user's training action on each frame image in the training course video frame sequence. The completion parameters of the action in each frame image in the training course video frame sequence can be understood as the data parameters that the user's limbs need to reach when the user completes the action in each frame image in the training course video frame sequence. The completion parameters of the standard training action completed by the user can be understood as the data parameters that the user's limbs need to reach when the user completes the standard training action according to the proportional parameters corresponding to the training action in the standard training action library completed by the user. The position can be determined by using a 6-axis sensor, a spatial positioning algorithm, and a depth camera or an RGB camera.
[0059] In an embodiment of the present invention, the height ratio, weight ratio, limb ratio, active joint position, age difference and gender are determined based on the trainee's physical sign information and the user's physical sign information in the standard training action library, and the proportion parameters corresponding to the training actions in the standard training action library are determined. The standard training actions corresponding to each frame in the training course video frame sequence are determined by matching the action features of each frame in the training course video frame sequence with the action features of the standard training actions in the standard training action library, and the completion parameters of the standard training actions completed by the user in the standard training action library are determined based on the standard training actions and the proportion parameters corresponding to the training actions in the standard training action library completed by the user. The completion rate of the actions completed by the user in each frame in the training course video frame sequence is determined based on the completion parameters of the matched standard training actions completed by the user and the completion parameters of the actions in each frame in the training course video frame sequence.
[0060] Furthermore, the determining of the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information includes:
[0061] Based on the height, weight, limb proportions, and active joints in the trainee's physical information in the standard training action library and the height, weight, limb proportions, and active joints in the user's physical information, the proportion parameters corresponding to the training actions in the standard training action library completed by the user are determined.
[0062] In the embodiment of the present invention, the height ratio, weight ratio, limb ratio, active joint position, age difference and gender are determined according to the trainee's physical sign information and the user's physical sign information in the standard training action library, and the proportion parameters corresponding to the training actions in the standard training action library are determined. For example, the ratio of the range of movement of the active joint position determined according to the user's limb ratio to the range of movement of the active joint position determined according to the trainee's limb ratio is used to determine the position ratio parameters corresponding to the user's limb ratio and the active joint point.
[0063] Furthermore, the determining of the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library includes:
[0064] Determining, according to the training course video frame sequence, the action features of each frame image in the training course video frame sequence and the completion parameters of the user action;
[0065] If a corresponding standard training action is matched with the action features of a standard training action in the standard training action library according to the action features of each frame image in the training course video frame sequence, then the completion parameters of the user completing the matched standard training action are determined according to the completion parameters of the corresponding standard training action, the proportion parameters corresponding to the training action in the standard training action library completed by the user, and the user's vital sign information.
[0066] In an embodiment of the present invention, the completion parameters of the user action can be understood as the completion parameters corresponding to the action completed by the user on each frame image in the training course video frame sequence, for example: the distance between the user's feet and the horizontal ground in each frame image in the training course video frame sequence, the spatial position of each active joint point, and the spatial angle formed by the limbs connected by each active joint point.
[0067] In the embodiment of the present invention, the action features of each frame image in the training course video frame sequence are identified through the training course video frame sequence, and the standard training action corresponding to the action features of each frame image in the training course video frame sequence is matched in the standard training action library according to the action features of each frame image in the training course video frame sequence. If the standard training action corresponding to the action features of each frame image in the training course video frame sequence is matched in the standard training action library according to the action features of each frame image in the training course video frame sequence, the completion parameters of the training action completed by the user are calculated according to the completion parameters of the standard training action, the proportion parameters corresponding to the training action completed by the user in the standard training action library, and the user's physical sign information. If the standard training action corresponding to the action features of each frame image in the training course video frame sequence cannot be matched in the standard training library according to the action features of each frame image in the training course video frame sequence, the corresponding training action is continued to be matched in the training action database according to the action features of each frame image in the training course video frame sequence until a matching training action is found, and the above-mentioned operation in the standard training library is repeated to calculate the completion parameters of the training action completed by the user.
[0068] The present invention obtains the action completion rate of each frame image in the training course video frame sequence according to user vital sign information, the user's training course video frame sequence and a standard training action library; the user vital sign information, two adjacent frames in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame image changing to the next frame image in the two adjacent frames; the calorie estimation model is obtained by training a deep learning network model through sample user vital sign information, two adjacent frames, the action completion rate corresponding to the two adjacent frames, and the calories consumed by the training action on the previous frame image changing to the next frame image in the two adjacent frames in the sample training course video frame sequence; and the total consumed calories of the training course video frame sequence are obtained according to the calories consumed by the training action on the previous frame image changing to the next frame image in the training course video frame sequence except the first frame image. This solution solves the problem that smart training plans and self-edited courses cannot estimate calories when there is no heart rate collection device, and achieves accurate estimation of the calorie value corresponding to the user's training course based on the user's physical sign information and the training course video frame sequence.
[0069] Embodiment 2
[0070] Figure 2 : is a flow chart of a calorie estimation method in Embodiment 2 of the present invention. The technical solution of the embodiment of the present invention is further refined on the basis of the above technical solution, and specifically mainly includes the following steps:
[0071] Step 210, obtaining a sample training course video frame sequence, training data corresponding to the sample training course video frame sequence, and sample user vital sign information in the sample training course video frame sequence;
[0072] In an embodiment of the present invention, the sample training course video frame sequence can be understood as the video frame sequence of the sample training course split when input when training the calorie estimation model, and can also be understood as the video frame sequence of the historical training course split in the training video library. The training data corresponding to the sample training course video frame sequence can be understood as the cumulative calorie value corresponding to the time point where each frame image in the sample training course video frame sequence is located, and the time point and time period corresponding to the intermittent rest time in the sample training course video frame sequence. The sample user's physical sign information in the sample training course video frame sequence can be understood as the physical sign information of the training user in the sample training course video frame sequence, such as: the training user's height, age, weight, gender, limb length, limb proportions, active joints and other related information.
[0073] In an embodiment of the present invention, a sample training course video is obtained from a historical training video library and the obtained sample course video is split into a sample training course video frame sequence, and training data corresponding to the sample training course video frame sequence and sample user vital signs information in the sample training course video frame sequence are simultaneously obtained to calculate the action completion rate of each frame image in the sample training course video frame sequence.
[0074] In the embodiment of the present invention, the intermittent rest time in the video frame sequence split from the sample training course can be understood as a video frame sequence that can be ignored or a video frame sequence image without a user screen in the video frame sequence. The intermittent rest time can also be understood as a video frame sequence segmentation point, and the complete video frame sequence is divided into two parts for related calorie calculation and finally accumulated into the total calories consumed by the entire video frame sequence.
[0075] Step 220, determining the action completion rate of each frame image in the sample training course video frame sequence according to the action features of each frame image in the sample training course video frame sequence, the action features of the training actions in the standard training action library, the sample user's physical sign information in the sample training course video frame sequence, and the trainee's physical sign information in the standard training library;
[0076] In the embodiment of the present invention, the action feature of each frame image in the sample training course video frame sequence can be understood as the action attribute information of the user's training action on each frame image in the sample training course video frame sequence. The action completion rate of each frame image in the sample training course video frame sequence can be understood as the completion percentage of the matching action in the standard training action library calculated based on the sample user's physical sign information and the trainer's physical sign information.
[0077] In the embodiment of the present invention, the action of each frame image in the sample user's training course video frame sequence is determined based on the sample user's physical sign information and the sample user's sample training course video frame sequence. The completion parameters corresponding to the training actions in the standard training action library completed by the sample user are determined based on the sample user's physical sign information and the trainer information in the standard training action library; the action completion rate of each frame image in the sample user's training course video frame sequence is determined based on the action completion parameters of each frame image in the sample user's training course video frame sequence and the completion parameters corresponding to the training actions in the standard training action library completed by the sample user.
[0078] Step 230, determining the calories consumed by changing the training action on the previous frame image to the training action on the next frame image in two adjacent frames in the sample training course video frame sequence according to the training data corresponding to the sample training course video frame sequence and the sample training course video frame sequence;
[0079] In an embodiment of the present invention, the calories consumed by the change from the training action on the previous frame to the training action on the next frame in two adjacent frames of the sample training course video frame sequence can be understood as the calorie value consumed by the sample user when the training action on the previous frame changes to the training action on the next frame.
[0080] In an embodiment of the present invention, the accumulated calorie value corresponding to the time point of each frame image in the sample training course video frame sequence in the training data corresponding to the sample training course video frame sequence is used, and the accumulated calorie value corresponding to the time point of the latter frame image in two adjacent frames in the sample training course video frame sequence is subtracted from the accumulated calorie value corresponding to the time point of the previous frame image, to obtain the calories consumed by the training action on the previous frame image in two adjacent frames in the training course video frame sequence.
[0081] Step 240, input the two adjacent frames of images in the sample training course video frame sequence, the sample user's vital signs information in the sample training course video frame sequence, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame image to the next frame image in the two adjacent frames of images in the sample training course video frame sequence into the deep learning network model to obtain a calorie estimation model.
[0082] In the embodiment of the present invention, the deep learning network model can be understood as a machine learning algorithm that can learn the inherent laws and representation levels of sample data. The deep learning network model can adopt, but is not limited to, machine learning models such as linear regression, polynomial regression, ridge regression, and neural network.
[0083] In an embodiment of the present invention, two adjacent frames of images in a sample training course video frame sequence, sample user vital signs information in the sample training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame image changing to the next frame image in the two adjacent frames of images in the sample training course video frame sequence are input into a deep learning network model, and then the deep learning network model analyzes the internal rules and characteristics of the two adjacent frames of images in the sample training course video frame sequence, the sample user vital signs information in the sample training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images in the sample training course video frame sequence according to the input adjacent two frames of images in the sample training course video frame sequence, the sample user vital signs information in the sample training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images, and performs training to obtain a calorie estimation model.
[0084] Optionally, a user calorie estimation model is obtained by inputting a deep learning network model based on a sample user's physical fitness training course video frame sequence, training data corresponding to the sample user's physical fitness training course video frame sequence, and sample user's vital sign information.
[0085] Optionally, a simple calorie estimation model is obtained by inputting a deep learning network model based on a sample user training course video frame sequence, training data corresponding to the sample user training course video frame sequence, and sample user vital sign information.
[0086] In the embodiment of the present invention, the data that can be input can be calculated based on the data obtained under the current situation, and a suitable estimation model can be selected from the calorie estimation model, the user calorie estimation model and the simple calorie estimation model to perform calorie estimation.
[0087] Furthermore, before obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the method further includes:
[0088] If the action feature of the first frame image in the training course video frame sequence matches the training start action in the standard training action library, the calories corresponding to the training start action in the standard training library are counted into the total consumed calories of the training course video frame sequence.
[0089] In the embodiment of the present invention, the training start action in the standard training library can be understood as the start action corresponding to the standard training course. The calories corresponding to the training start action in the standard training library can be understood as the calorie consumption corresponding to the training start action of the trainee entering the standard training library from the natural standing action.
[0090] In an embodiment of the present invention, when the total calories consumed by a training course video frame sequence is used as the total calories consumed when a user trains the training course, it is necessary to match the motion features of the training action on the first frame image in the training course video frame sequence with the motion features of the training start action in the standard training library. If the motion features of the first frame image in the training course video frame sequence match the training start action in the standard training action library, the calories corresponding to the training start action in the standard training library are included in the total calories consumed by the training course video frame sequence.
[0091] Furthermore, before obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the method further includes:
[0092] If the action features of the first frame image in the training course video frame sequence match the training start action in the standard training action library, the calories corresponding to the matching action in the standard training action library based on the action features of the first frame image in the training course video frame sequence will be counted into the total consumed calories of the training course video frame sequence.
[0093] In an embodiment of the present invention, when the total calories consumed by a training course video frame sequence is used as the total calories consumed when a user trains a training course, it is necessary to match the motion features of the training action on the first frame image in the training course video frame sequence with the motion features of the training start action in the standard training library. If the motion features of the first frame image in the training course video frame sequence do not match the training start action in the standard training action library, the calories corresponding to the matching action in the standard training action library according to the motion features of the first frame image in the training course video frame sequence are counted into the total calories consumed by the training course video frame sequence. Among them, the calories corresponding to the matching action can be understood as the calories consumed by the trainee from the natural standing action to the training matching action in the standard training library.
[0094] The present invention obtains a sample training course video frame sequence, training data corresponding to the sample training course video frame sequence, and sample user vital signs information in the sample training course video frame sequence; determines the action completion rate of each frame image in the sample training course video frame sequence according to the action features of each frame image in the sample training course video frame sequence, the action features of the training actions in the standard training action library, the sample user vital signs information in the sample training course video frame sequence, and the trainer feature information in the standard training library; determines the sample training course video frame sequence according to the training data corresponding to the sample training course video frame sequence and the sample training course video frame sequence. The calories consumed by the training action on the previous frame image of two adjacent frames in the video frame sequence are changed to the training action on the next frame image; the user's physical sign information, the adjacent two frames in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames are input into the deep learning network model to obtain a calorie estimation model. The problem that the smart training plan and self-compiled courses cannot perform calorie estimation in the absence of a heart rate acquisition device is solved, and the calorie value corresponding to the user's training course is accurately estimated based on the user's physical sign information and the training course video frame sequence.
[0095] Embodiment 3
[0096] Figure 3 A calorie estimation device provided by an embodiment of the present invention includes: a completion rate determination module 310, a calorie determination module 320 and a total consumption confirmation module 330;
[0097] The completion rate determination module 310 is used to obtain the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library;
[0098] The calorie determination module 320 is used to input the user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through the sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images;
[0099] The total consumption confirmation module 330 is used to obtain the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence.
[0100] Furthermore, the completion rate determination 310 module is specifically used for:
[0101] Determine the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information;
[0102] Determine the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library;
[0103] The completion rate of the user's completed action in each frame in the training course video frame sequence is determined according to the completion parameters of the user's completed matching standard training action and the completion parameters of the action in each frame in the training course video frame sequence.
[0104] Furthermore, the completion rate determination 310 module is specifically used for:
[0105] Based on the height, weight, limb proportions, and active joints in the trainee's physical information in the standard training action library and the height, weight, limb proportions, and active joints in the user's physical information, the proportion parameters corresponding to the training actions in the standard training action library completed by the user are determined.
[0106] Furthermore, the completion rate determination module 310 is specifically used for:
[0107] Determining, according to the training course video frame sequence, the action features of each frame image in the training course video frame sequence and the completion parameters of the user action;
[0108] If a corresponding standard training action is matched with the action features of a standard training action in the standard training action library according to the action features of each frame image in the training course video frame sequence, then the completion parameters of the user completing the matched standard training action are determined according to the completion parameters of the corresponding standard training action, the proportion parameters corresponding to the training action in the standard training action library completed by the user, and the user's vital sign information.
[0109] Furthermore, the calorie determination module 320 is further configured to:
[0110] Acquire a sample training course video frame sequence, training data corresponding to the sample training course video frame sequence, and sample user vital sign information in the sample training course video frame sequence;
[0111] Determine the action completion rate of each frame image in the sample training course video frame sequence according to the action features of each frame image in the sample training course video frame sequence, the action features of the training actions in the standard training action library, the sample user's physical sign information in the sample training course video frame sequence and the trainer feature information in the standard training library;
[0112] Determine, based on the training data corresponding to the sample training course video frame sequence and the sample training course video frame sequence, the calories consumed by changing from a training action on a previous frame image to a training action on a subsequent frame image in two adjacent frames of the sample training course video frame sequence;
[0113] The two adjacent frames of images in the sample training course video frame sequence, the sample user's vital signs information in the sample training course video frame sequence, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame image changed to the training action on the next frame image in the two adjacent frames of images in the sample training course video frame sequence are input into the deep learning network model to obtain a calorie estimation model.
[0114] Furthermore, the total consumption confirmation module 330 is further used to:
[0115] If the action feature of the first frame image in the training course video frame sequence matches the training start action in the standard training action library, the calories corresponding to the training start action in the standard training library are counted into the total consumed calories of the training course video frame sequence.
[0116] Furthermore, the total consumption confirmation module 330 is further used to:
[0117] If the action features of the first frame image in the training course video frame sequence match the training start action in the standard training action library, the calories corresponding to the matching action in the standard training action library based on the action features of the first frame image in the training course video frame sequence will be counted into the total consumed calories of the training course video frame sequence.
[0118] A calorie estimation device provided in an embodiment of the present invention can execute a calorie estimation method provided in any embodiment of the present invention, and has corresponding functional modules and beneficial effects of the execution method.
[0119] Embodiment 4
[0120] Figure 4 A schematic diagram of the structure of an electronic device provided in Example 4 of the present invention. Figure 4 A block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the present invention is shown. Figure 4 The electronic device 12 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.
[0121] like Figure 4 As shown, the electronic device 12 is in the form of a general purpose computing device. The components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that connects various system components (including the system memory 28 and the processing unit 16).
[0122] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.
[0123] The electronic device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.
[0124] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.
[0125] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in the memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 42 generally perform the functions and / or methods of the embodiments described herein.
[0126] The electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), may communicate with one or more devices that enable a user to interact with the electronic device 12, and / or may communicate with any device that enables the electronic device 12 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the electronic device 12 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via a bus 18. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0127] The processing unit 16 executes various functional applications and data processing by running the program stored in the system memory 28, for example, implementing a calorie estimation method provided by an embodiment of the present invention, the method comprising:
[0128] According to the user's physical sign information, the user's training course video frame sequence and the standard training action library, the action completion rate of each frame image in the training course video frame sequence is obtained;
[0129] The user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through the sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images;
[0130] According to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the total consumed calories of the training course video frame sequence are obtained.
[0131] Embodiment 5
[0132] Embodiment 5 of the present invention further provides a storage medium including computer executable instructions, wherein the computer executable instructions are used to perform a calorie estimation method when executed by a computer processor, the method comprising:
[0133] According to the user's physical sign information, the user's training course video frame sequence and the standard training action library, the action completion rate of each frame image in the training course video frame sequence is obtained;
[0134] The user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through the sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images;
[0135] According to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the total consumed calories of the training course video frame sequence are obtained.
[0136] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, 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), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, a device or a device or used in combination with it.
[0137] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0138] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0139] Computer program code for performing the operations of the present invention may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).
[0140] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A calorie estimation method, characterized in that: include: According to the user's physical sign information, the user's training course video frame sequence and the standard training action library, the action completion rate of each frame image in the training course video frame sequence is obtained; The user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images are input into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through the sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; According to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the total consumed calories of the training course video frame sequence are obtained; The step of obtaining the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library includes: Determine the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information; Determine the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library; The completion rate of the user's action in each frame in the training course video frame sequence is determined according to the completion parameters of the user's completed matching standard training action and the completion parameters of the action in each frame in the training course video frame sequence.
2. The method according to claim 1, characterized in that The determining, based on the trainee's physical sign information in the standard training action library and the user's physical sign information, a proportion parameter corresponding to the user completing the training action in the standard training action library comprises: Based on the height, weight, limb proportions, and active joints in the trainee's physical information in the standard training action library and the height, weight, limb proportions, and active joints in the user's physical information, the proportional parameters corresponding to the training actions in the standard training action library completed by the user are determined.
3. The method according to claim 1, characterized in that: The determining, based on the action features of each frame image in the training course video frame sequence, the action features of the standard training actions in the standard training action library, and the proportion parameters corresponding to the user completing the training actions in the standard training action library, the completion parameters of the user completing the matching standard training action include: Determining, according to the training course video frame sequence, the action features of each frame image in the training course video frame sequence and the completion parameters of the user action; If a corresponding standard training action is matched with the action features of a standard training action in the standard training action library according to the action features of each frame image in the training course video frame sequence, then the completion parameters of the user completing the matched standard training action are determined according to the completion parameters of the corresponding standard training action, the proportion parameters corresponding to the training action in the standard training action library completed by the user, and the user's vital sign information.
4. The method according to claim 1, characterized in that: The calorie estimation model is obtained by training a deep learning network model through sample user vital signs information in a sample training course video frame sequence, two adjacent frames of images, action completion rates corresponding to the two adjacent frames of images, and calories consumed by the training action on the previous frame of the two adjacent frames of images. The model includes: Acquire a sample training course video frame sequence, training data corresponding to the sample training course video frame sequence, and sample user vital sign information in the sample training course video frame sequence; Determine the action completion rate of each frame image in the sample training course video frame sequence according to the action features of each frame image in the sample training course video frame sequence, the action features of the training actions in the standard training action library, the sample user's physical sign information in the sample training course video frame sequence and the trainer feature information in the standard training library; Determine, based on the training data corresponding to the sample training course video frame sequence and the sample training course video frame sequence, the calories consumed by changing from a training action on a previous frame image to a training action on a subsequent frame image in two adjacent frames of the sample training course video frame sequence; The two adjacent frames of images in the sample training course video frame sequence, the sample user's vital signs information in the sample training course video frame sequence, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame image changed to the training action on the next frame image in the two adjacent frames of images in the sample training course video frame sequence are input into the deep learning network model to obtain a calorie estimation model.
5. The method according to claim 1, characterized in that Before obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the method further includes: If the action feature of the first frame image in the training course video frame sequence matches the training start action in the standard training action library, the calories corresponding to the training start action in the standard training library are counted into the total consumed calories of the training course video frame sequence.
6. The method according to claim 1, characterized in that Before obtaining the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence, the method further includes: If the action features of the first frame image in the training course video frame sequence match the training start action in the standard training action library, the calories corresponding to the matching action in the standard training action library based on the action features of the first frame image in the training course video frame sequence will be counted into the total consumed calories of the training course video frame sequence.
7. A calorie estimation device, characterized in that: include: A completion rate determination module, used to obtain the action completion rate of each frame image in the training course video frame sequence according to the user's physical sign information, the user's training course video frame sequence and the standard training action library; A calorie determination module is used to input the user's physical sign information, two adjacent frames of images in the training course video frame sequence, and the action completion rate corresponding to the two adjacent frames of images into a calorie estimation model to obtain the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; the calorie estimation model is obtained by training a deep learning network model through sample user's physical sign information, two adjacent frames of images, the action completion rate corresponding to the two adjacent frames of images, and the calories consumed by the training action on the previous frame of the two adjacent frames of images changing to the training action on the next frame of the two adjacent frames of images; A total consumption confirmation module is used to obtain the total consumed calories of the training course video frame sequence according to the calories consumed by the training action on the previous frame image except the first frame image in the training course video frame sequence; The completion rate determination module is specifically used for: Determine the proportion parameter corresponding to the user completing the training action in the standard training action library according to the trainee's physical sign information in the standard training action library and the user's physical sign information; Determine the completion parameters of the user completing the matching standard training action according to the action features of each frame image in the training course video frame sequence, the action features of the standard training action in the standard training action library, and the proportion parameters corresponding to the user completing the training action in the standard training action library; The completion rate of the user's action in each frame in the training course video frame sequence is determined according to the completion parameters of the user's completed matching standard training action and the completion parameters of the action in each frame in the training course video frame sequence.
8. An electronic device, characterized in that: The electronic device comprises: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the calorie estimation method according to any one of claims 1 to 6.
9. A storage medium comprising computer executable instructions, wherein the computer executable instructions are used to perform the calorie estimation method according to any one of claims 1 to 6 when executed by a computer processor.
