KR20200083104A - Dance training apparatus using automatic difficulty control of dance motion - Google Patents

Abstract

Disclosed is a dance training apparatus using automatic difficulty control of dance motion, which comprises: an animation unit (100) making 3D dance animation of each level; a storage unit (200) storing the 3D dance animation; and a providing unit (300) providing the 3D dance animation by each level. According to the present invention, since important motion video is automatically generated and provided when dance teaching content is generated, a large number of teaching content can be made quickly. While a method in which a user learns dance motion while viewing video is one-directional, the present invention provides the user with a matching ratio of each portion with respect to the important motion, such that a weak portion is recognized by each dance segment and repetitive learning is provided to each dance section to enable the user to quickly master the dance motion. Moreover, beginners widely set an interval in the automatically generated and divided important motion video to learn large motion first and shortly set the interval of the same after being skilled by a predetermined degree to learn detailed motion, thereby providing customized learning by a user′s level.

Description

Dance training device using automatic difficulty control of dance motion{DANCE TRAINING APPARATUS USING AUTOMATIC DIFFICULTY CONTROL OF DANCE MOTION}

The present invention relates to a dance training device using automatic difficulty adjustment of a dance action, and more particularly, to a dance training device using automatic difficulty adjustment of a dance action that provides a dance video.

The dance game started with Konami's Dance Dance Revolution (DDR), in which a symbol displayed on the scaffold, a hardware device, was displayed on the screen, and it was followed in succession to induce similar behavior to dance. However, this dance movement is a dance movement created by moving the sign of the footing regardless of the dance of the corresponding music.

K-pop is becoming popular all over the world, and the cover dance culture that follows the idol group's dance is born as a new Korean wave, and the cover dance videos are shared on YouTube and other video sharing websites worldwide.

Existing ways to learn cover dance exist by watching videos from these video sharing sites and following them, or by receiving lessons from dance academies.

In the Xbox360, Microsoft's game console, Dance Central using Kinect, a motion camera sensor, is provided as a game that can dance based on the user's movement, but focused on gameplay rather than the actual dance class.

Registration number: 10-0457201, dance game device

An object of the present invention for solving the above problems is to provide a dance training apparatus using automatic difficulty control of a dance motion that provides a 3D dance animation that a user can follow according to the difficulty level.

The present invention for achieving the above object, the animation unit 100 to create a 3D dance animation for each level; A storage unit 200 for storing 3D dance animations; And a providing unit 300 providing 3D dance animation for each level.

In addition, the animation unit 100 includes a motion comparison unit 110 for comparing the inter-frame motion of 3D spatial data; A frame selector 120 for selecting a key motion frame at a specific time interval or a change rate of a dance motion according to the motion comparison result of the motion comparison unit 110; An interpolation unit 130 interpolating animation between selected key motion frames; And a post-processing unit 140 that outputs a 3D dance animation that is interpolated by post-processing through an editing tool.

In addition, the providing unit 300 provides a 2D image key motion snapshot in 3D dance animation.

In addition, it further includes an evaluation unit 400 for evaluating dance training based on 3D dance animation and camera images, and the frame selection unit 120 provides feedback to the neural network 121 with feedback from the providing unit 300 and the evaluation unit 400. By selecting the key motion frame.

In addition, the frame selector 120 includes a neural network 121 for selecting a key motion frame; And a feedback unit 122 feeding back the outputs of the providing unit 300 and the evaluation unit 400 to the neural network 121, the neural network 121 receiving a 3D dance animation and a camera image and providing the providing unit 300. The key motion frame is selected according to the input of the and evaluation unit 400, and the neural network of the neural network 121 is a 3D dance animation, a camera image, and a key motion frame according to the input of the providing unit 300 and the evaluation unit 400. A plurality of multipliers (1211) for selecting, learning the feedback according to the output of the providing unit 300 and the evaluation unit 400 according to the key motion frame, and outputting a value multiplied by the weight of the 3D dance animation input; An adder 1212 for adding a plurality of multiplier 1211 outputs and outputting a value for selecting a key motion frame in a 3D dance animation; And a weight setting unit 1213 that is set by the feedback unit 122 to set weights input to the multipliers 1211, and the weight setting unit 1213 selects different weight settings according to the difficulty level. It includes a level selector.

In addition, the level selection unit, the lower level selection unit for setting the first weight to a plurality of multipliers 1211 when the level of difficulty; A level selection unit for setting a second weight to a plurality of multipliers 1211 when the level is in the difficulty level; And a phase level selector configured to set the third weight value in the multiplier 1211 when the difficulty level is high.

In the case of using the dance training apparatus using the automatic difficulty control of the dance motion according to the present invention as described above, in generating dance teaching content, it is possible to quickly create and present many teaching content by automatically generating and presenting important motions.

Also, While the way of learning while watching a video is unidirectional, the present invention recognizes vulnerable areas for each dance section by presenting a matching rate for each part to the important motion to the user, and masters the dance motion quickly by suggesting repetitive learning about the vulnerable areas. Can.

In addition, beginners can learn the big action first by widening the interval of the important classification action that is automatically generated, and if you are somewhat skilled, you can learn the detailed action by shortening the interval of the important classification action to learn the detailed action. Do.

1 is a schematic diagram of a method for automatically adjusting the difficulty of dance.
2 is an exemplary view showing a dance training method.
3 is a block diagram of a dance training device.

The terms used in the specification can be interpreted by expanding the meanings of the original words, and the words and words having different meanings are combined into a single word, thus taking the form of a compound word having a new meaning in a broad sense. For example, the word dance and the word training are combined to achieve dance training, which means that dance training is performed. In addition, the embodiments presented in the specification are preferred embodiments for implementation, and it is possible that other components are added or original components are omitted depending on the case. In embodiments, it may be implemented with or without configurations.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The present invention is a dance training device that allows a user to practice dance with a virtual 3D dance trainer on the screen by himself.

The dance training device automatically divides the motion section to learn a dance composed of fast continuous motions, and omits continuous motions that are difficult for beginners to follow. It provides a function to gradually learn the skipped continuous motion.

In particular, the dance training device repeatedly presents vulnerable sections indicated by a low match rate, and learns how to increase the match rate by letting the user know where the match rate is low for each part of the body and continuously repeat it so that the user can continuously follow. To suggest.

In the prior art, a key motion snapshot in the form of a 2D image is automatically generated and presented on a screen, and a method for inferring or following a dance motion through the corresponding image is proposed.

In the present invention, the key motion of the 3D animation is selected, the animation of the difficult continuous motion is boldly omitted, and the omitted animation is automatically generated through interpolation (interpolation), so that the dance motion similar to the original dance motion is easy to learn. Here's how to do it.

1 is a schematic diagram of a method for automatically adjusting the difficulty of dance.

The dance training device compares the inter-frame motion of the 3D spatial data (3D animation data) of the dance motion.

The dance training device selects a key motion frame at a specific time interval or a rate of change in dance movement. At this time, if the rate of change and the time interval are held low, more frames are selected, more similar to the original dance motion, and the difficulty of the dance motion increases.

Animation between selected key motion frames is automatically generated through interpolation. Animation interpolation is a function commonly provided by general commercial 3D game engines (eg Unity 3D, Unreal Engine, etc.).

The generated 3D animation is reviewed through an editing tool, and post-processing is performed to correct awkward parts. At this time, considering the parameters such as frame interval or time interval, the difficulty is classified and stored as beginner, intermediate, advanced, and expert. The advanced or expert grade can be the same as the original dance moves.

The dance training device stores 3D dance animations for each difficulty level, and provides dance movements of a suitable difficulty level through user selection or real-time level evaluation of the user.

2 is an exemplary view showing a dance training method.

The dance training device selects a key motion frame according to the difficulty of dance from the original dance motion for the professional, and provides 3D dance animation, which is a generated low-level beginner dance motion. The original dance moves for professionals are high-level and advanced dance moves that have dance moves that are not easy for users to follow from the beginning. The dance training device extracts a dance motion that is easy for a user to follow from a professional dance motion original and provides a low-level beginner dance motion 3D dance animation. The dance training device selects a key motion frame from the original dance motion for professional use to provide 3D dance animation. When the dance training device selects a key motion frame, it is possible to compare the motions between frames and select a key motion frame at a change rate or a specific time interval of the dance motion. The dance training device for creating a 3D dance animation by selecting a key motion frame is as follows.

3 is a block diagram of a dance training device.

The dance training apparatus includes an animation unit 100 that creates 3D dance animations for each level; A storage unit 200 for storing 3D dance animations; And a providing unit 300 providing 3D dance animation for each level.

The animation unit 100 includes an operation comparison unit 110 for comparing inter-frame operations of 3D spatial data; A frame selector 120 for selecting a key motion frame at a specific time interval or a change rate of a dance motion according to the motion comparison result of the motion comparison unit 110; An interpolation unit 130 interpolating animation between selected key motion frames; And a post-processing unit 140 that outputs a 3D dance animation that is interpolated by post-processing through an editing tool.

The providing unit 300 provides a 2D image key motion snapshot in 3D dance animation.

Further comprising an evaluation unit 400 for evaluating dance training based on 3D dance animation and camera images, the frame selection unit 120 is keyed by the neural network 121 with feedback from the providing unit 300 and the evaluation unit 400. Select a motion frame.

The frame selector 120 includes a neural network 121 for selecting a key motion frame; And a feedback unit 122 feeding back the outputs of the providing unit 300 and the evaluation unit 400 to the neural network 121, the neural network 121 receiving a 3D dance animation and a camera image, and providing the providing unit 300 The key motion frame is selected according to the input of the and evaluation unit 400, and the neural network of the neural network 121 is a 3D dance animation, a camera image, and a key motion frame according to the input of the providing unit 300 and the evaluation unit 400. A plurality of multipliers (1211) for selecting, and learning the feedback according to the output of the providing unit 300 and the evaluation unit 400 according to the key motion frame, and outputting a value multiplied by the weight of the 3D dance animation input; A multiplier 1212 that outputs a value for selecting a key motion frame in a 3D dance animation by adding multiple multiplier 1211 outputs; And a weight setting unit 1213 that is set by the feedback unit 122 to set weights input to the multipliers 1211, and the weight setting unit 1213 selects different weight settings according to the difficulty level. It includes a level selector.

The level selector is a lower level selector for setting a first weight to a plurality of multipliers 1211 when the difficulty level is lower; A level selection unit for setting a second weight to a plurality of multipliers 1211 when the level is in the difficulty level; And a phase level selector configured to set the third weight value in the multiplier 1211 when the difficulty level is high.

The level selection unit sets a weight according to the difficulty level, and thus represents the effect that the dance training device can generate a 3D dance animation in which the difficulty level changes.

Although described above with reference to preferred embodiments of the present invention, those skilled in the art variously modify and change the present invention without departing from the spirit and scope of the present invention as set forth in the claims below. You can understand that you can.

100: animation unit 110: motion comparison unit
120: frame selector 121: neural network
1211: multiplier 1212: adder
1213: weight setting unit 122: feedback unit
130: interpolation unit 140: post-processing unit
200: storage unit 300: providing unit
400: evaluation unit

Claims (5)

Animation unit 100 to create a 3D dance animation for each level;
A storage unit 200 for storing 3D dance animations; And
Dance training device using the automatic difficulty adjustment of the dance movement, including a providing unit 300 for providing 3D dance animation for each level. According to claim 1,
The animation unit 100 includes an operation comparison unit 110 for comparing inter-frame operation of 3D spatial data;
A frame selector 120 for selecting a key motion frame at a specific time interval or a change rate of a dance motion according to the motion comparison result of the motion comparison unit 110;
An interpolation unit 130 interpolating animation between selected key motion frames; And
Dance training device using the automatic difficulty control of the dance movement, including a post-processing unit 140 for outputting a 3D dance animation that is interpolated by post-processing through an editing tool. According to claim 1,
The providing unit 300 is a dance training apparatus using automatic difficulty adjustment of a dance motion that provides a 2D image key motion snapshot in a 3D dance animation. According to claim 2,
Further comprising an evaluation unit 400 for evaluating dance training by 3D dance animation and camera video,
The frame selection unit 120 is a dance training apparatus using automatic difficulty control of a dance operation to select a key motion frame by the neural network 121 with feedback from the providing unit 300 and the evaluation unit 400. According to claim 4,
The frame selector 120 includes a neural network 121 for selecting a key motion frame; And
Dance training device using the automatic difficulty control of the dance operation, including a feedback unit (122) for feeding back the output of the providing unit (300) and the evaluation unit (400) to the neural network (121).

Images