CN107092894A - A kind of motor behavior recognition methods based on LSTM models - Google Patents
A kind of motor behavior recognition methods based on LSTM models Download PDFInfo
- Publication number
- CN107092894A CN107092894A CN201710292408.4A CN201710292408A CN107092894A CN 107092894 A CN107092894 A CN 107092894A CN 201710292408 A CN201710292408 A CN 201710292408A CN 107092894 A CN107092894 A CN 107092894A
- Authority
- CN
- China
- Prior art keywords
- data
- lstm
- motor behavior
- recognition methods
- motion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The present invention solves the technical problem of provide a kind of motor behavior recognition methods based on LSTM models, good recognition accuracy can be obtained with relatively small number of data, overcoming simultaneously needs the artificial deficiency for extracting feature in current class algorithm, enable extensive use in practice.Step S1:By being worn over the time series data that the sensor bracelet of human hand collects relevant multigroup motion of different people;Step S2:Its corresponding type of sports is marked out to the multidimensional initial data collected, it is ensured that the progress of following supervised learning;Step S3:Necessary processing is carried out to training data, then LSTM pattern types is passed to as input data and is trained, optimal neural network parameter is obtained, is used as final identification model;Step S4:Motor behavior data to be identified are pre-processed, and as the input of LSTM pattern types, calculates the motion sequence of output layer maximum probability, regard the result as the sports category finally known.
Description
Technical field
Field is recognized the present invention relates to the human motion in general fit calculation, is in particular a kind of based on LSTM models
Motor behavior recognition methods.
Background technology
General fit calculation be also known as it is general deposit calculating, popularization calculate, this concept emphasizes the calculating combined together with environment, and calculates
Machine then disappears in the sight of people in itself.Under the pattern of general fit calculation, people can at any time and any place, with
Any mode enters the acquisition and processing of row information.Between connection breaking and light weight to calculate (i.e. computing resource is relatively limited) be general fit calculation
Most important two features.The software engineering of general fit calculation seeks to realize affairs and data processing in such a case.
The motion identification of early stage is mainly based upon visual manner, gives one section of image sequence or a video segment,
Identify the type of sports of personage.The method of view-based access control model has the advantages that interaction naturally, the characteristic information extracted enriches, but is somebody's turn to do
Method also has some limitations in actual applications, it is desirable to overcome many problems.Such as the illumination condition in environment, personage is in shooting
Position before machine, the size in place etc..
With the popularization of motion bracelet and intelligent watch, sensor-based motion identification becomes noticeable all the more.Pass
Sensor has cheap, easy to carry, the advantages of not limited by place, with the development of these equipment, motion identification and quilt
A piece of new research field is brought into, the deficiency of the motion recognition method supplemented with traditional view-based access control model in actual applications promotees
The application of motion identification is in daily life made.Prevailing technical method includes using body in motor behavior field
The sensor of wearing, manual designs feature extraction program, and various (supervision) sorting techniques.Traditional recognition method is generally required
It is artificial to extract feature, and different features are often extracted in different motions.Therefore inconvenience can be brought in actual applications.
And be above-mentioned present in tional identification because its requirement to initial data is less with the rise of deep neural network
Problem brings new resolving ideas.
The content of the invention
, can the present invention solves the technical problem of a kind of motor behavior recognition methods based on LSTM models is provided
Good recognition accuracy is obtained with relatively small number of data, while overcoming needs artificial extraction feature in current class algorithm
Deficiency, enables extensive use in practice.
In order to solve the above technical problems, recognizing field the present invention relates to the human motion in general fit calculation, in particular
It is a kind of motor behavior recognition methods based on LSTM models, this method comprises the following steps:
Step S1:By being worn over the time series that the sensor bracelet of human hand collects relevant multigroup motion of different people
Data;
Step S2:Its corresponding type of sports is marked out to the multidimensional initial data collected, it is ensured that next supervision is learned
The progress of habit;
Step S3:Necessary processing is carried out to training data, then LSTM pattern types are passed to as input data
It is trained, obtains optimal neural network parameter, be used as final identification model;
Step S4:Motor behavior data to be identified are pre-processed, and as the input of LSTM pattern types, meter
The motion sequence of output layer maximum probability is calculated, the result is regard as the sports category finally known.
It is used as the further optimization of the technical program, a kind of motor behavior recognition methods institute based on LSTM models of the present invention
The step S1 stated is specially:The time series data during motion of people is obtained with motion bracelet, including heart rate, 3-axis acceleration are passed
Sensor data.
It is used as the further optimization of the technical program, a kind of motor behavior recognition methods institute based on LSTM models of the present invention
The step S2 stated is specially:The data for the different test objects being collected into are labeled according to its sports category at that time, shape
It can be used for the complete data set that supervised learning dimension is F into one.
It is used as the further optimization of the technical program, a kind of motor behavior recognition methods institute based on LSTM models of the present invention
The step S3 stated is specially:The data being collected into are pre-processed as steps described below, remove transition state motion shape first
The data of state, fill missing values, remove time mark, then according to sensor sample frequency fHz, and one window size of design is
2f, step-length is split for f sliding window to time series data;Then the data split are normalized,
Comply with the form of sigmoid functions;Finally the incoming LSTM patterns type of the data handled well is trained, wherein losing letter
Number is categorical cross-entropy, and optimizer is RMSprop, and learning rate is 0.001;Obtained by training pattern
The corresponding weight parameter of different motion classification.
A kind of of the invention motor behavior recognition methods based on LSTM models has the beneficial effect that:
A kind of motor behavior recognition methods based on LSTM models of the present invention, can obtain good with relatively small number of data
Recognition accuracy, while overcoming needs the artificial deficiency for extracting feature in current class algorithm, enables wide in practice
General application.
Brief description of the drawings
The present invention will be further described in detail with specific implementation method below in conjunction with the accompanying drawings.
Fig. 1 is the present invention a kind of Data Collection, processing and the LSTM moulds of the motor behavior recognition methods based on LSTM models
The flow chart that type is built.
The LSTM illustratons of model that Fig. 2 builds for a kind of motor behavior recognition methods based on LSTM models of the present invention.
Fig. 3 is a kind of LSTM unit expanded schematic diagrams of the motor behavior recognition methods based on LSTM models of the present invention.
Fig. 4 is detailed annotation signal inside a kind of LSTM units of the motor behavior recognition methods based on LSTM models of the present invention
Figure.
Embodiment
For the object, technical solutions and advantages of the present invention are more clearly understood, below in conjunction with specific embodiment, and reference
Accompanying drawing, the present invention is described in more detail.
Step S1 is specially:The time series data during motion of people, specially heart rate monitor are obtained with motion bracelet
The acceleration on XYZ directions that heart rate, the 3-axis acceleration sensor being collected into are collected.
Step S2 is specially:The data for the different test objects being collected into are entered into rower according to its sports category at that time
Note, forms one and can be used for the complete data set that supervised learning dimension is F.
As shown in figure 1, the step S3 is specially:To the data that are collected into according to pre-processing.Remove transition first
The data of state motion state, fill missing values, remove time mark, then according to sensor sample frequency fHz, design one
Window size is 2f, and step-length is split for f sliding window to time series data.Then the data split are returned
One change is handled, and complies with the form of sigmoid functions.Finally the incoming designed LSTM patterns type of the data handled well is entered
Row training, wherein loss function are categorical cross-entropy, and optimizer is RMSprop, and learning rate is 0.001.
The corresponding weight parameter of different motion classification is obtained by training pattern.
As shown in Fig. 2 described LSTM patterns type include input layer, hidden layer, disconnection layer and output layer, input layer it is defeated
Enter the sliding window for being T for time step number, output layer is output as sports category sequence a1, a2, a3 ... am corresponding to input,
Activation primitive is softmax functions, and hidden layer includes multiple LSTM units, and the disconnection rate of disconnection layer is 50%, to prevent plan
Close.Time step is T, and dimension is F.
As shown in figure 3, the Temporal dependency in order to study mobile data, we used recursive data network, it is based on
The LSTM units of vanilla modifications.When some connections in network form directed circulation, the structure is recursive, wherein when
Preceding time t can take into account prior time t-1 network state.When the derivative of mistake passes through the plurality of layers in Recursive Networks
When " passage time " carries out backpropagation, LSTM units are used to suppress gradient decline.Each LSTM unit (joint) can be held
Continuous tracking represents the internal state of his " memory ".Over time, the unit association, exports, covering, or based on current
Output and past internal state empty their internal memory so that system retains the information of hundreds of time steps.
As shown in figure 4, LSTM units include 3 control doors (Input Gate, Output Gate, Forget gate),
For the association between control input, the internal state three for exporting and crossing over time step itself.Wherein ft=σ (Wf·
[ht-1, xt]+bf);it=σ (Wi·[ht-1, xt]+bi);
ot=σ (Wo[ht-1, xt]+bo);ht=ot*tanh(Ct)。
The step is specially that the test data come using collecting produces LSTM models finally to the time sequence to be identified
Row are identified, i.e., initial data is pre-processed first, become the data set that model can be recognized, then pass through mould
Type is predicted, to generate sports category result.
Certainly, described above not limitation of the present invention, the present invention is also not limited to the example above, the art
The variations, modifications, additions or substitutions that those of ordinary skill is made in the essential scope of the present invention, fall within the guarantor of the present invention
Protect scope.
Claims (4)
1. a kind of motor behavior recognition methods based on LSTM models, it is characterised in that this method comprises the following steps:
Step S1:By being worn over the time series number that the sensor bracelet of human hand collects relevant multigroup motion of different people
According to;
Step S2:Its corresponding type of sports is marked out to the multidimensional initial data collected, it is ensured that following supervised learning
Carry out;
Step S3:Necessary processing is carried out to training data, being then passed to LSTM patterns type as input data is carried out
Training, obtains optimal neural network parameter, is used as final identification model;
Step S4:Motor behavior data to be identified are pre-processed, and as the input of LSTM pattern types, calculated defeated
Go out the motion sequence of layer maximum probability, regard the result as the sports category finally known.
2. a kind of motor behavior recognition methods based on LSTM models according to claim 1, it is characterised in that:Described
Step S1 is specially:The time series data during motion of people, including heart rate, 3-axis acceleration sensor are obtained with motion bracelet
Data.
3. a kind of motor behavior recognition methods based on LSTM models according to claim 1, it is characterised in that:Described
Step S2 is specially:The data for the different test objects being collected into are labeled according to its sports category at that time, one is formed
It is individual to can be used for the complete data set that supervised learning dimension is F.
4. a kind of motor behavior recognition methods based on LSTM models according to claim 1, it is characterised in that:Described
Step S3 is specially:The data being collected into are pre-processed as steps described below, remove transition state motion state first
Data, fill missing values, remove time mark, then according to sensor sample frequency fHz, and one window size of design is 2f,
Step-length is split for f sliding window to time series data;Then the data split are normalized, make it
Meet the form of sigmoid functions;Finally the incoming LSTM patterns type of the data handled well is trained, wherein loss function is
Categorical cross-entropy, optimizer is RMSprop, and learning rate is 0.001;Obtain different by training pattern
The corresponding weight parameter of sports category.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710292408.4A CN107092894A (en) | 2017-04-28 | 2017-04-28 | A kind of motor behavior recognition methods based on LSTM models |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710292408.4A CN107092894A (en) | 2017-04-28 | 2017-04-28 | A kind of motor behavior recognition methods based on LSTM models |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107092894A true CN107092894A (en) | 2017-08-25 |
Family
ID=59637262
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710292408.4A Pending CN107092894A (en) | 2017-04-28 | 2017-04-28 | A kind of motor behavior recognition methods based on LSTM models |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107092894A (en) |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108926338A (en) * | 2018-05-31 | 2018-12-04 | 中南民族大学 | Heart rate prediction technique and device based on deep learning |
CN109214285A (en) * | 2018-08-01 | 2019-01-15 | 浙江深眸科技有限公司 | Detection method is fallen down based on depth convolutional neural networks and shot and long term memory network |
CN109344960A (en) * | 2018-09-01 | 2019-02-15 | 哈尔滨工程大学 | A kind of DGRU neural network and its prediction model method for building up preventing data information loss |
CN109447162A (en) * | 2018-11-01 | 2019-03-08 | 山东大学 | A kind of real-time Activity recognition system and its working method based on Lora and Capsule |
CN109447164A (en) * | 2018-11-01 | 2019-03-08 | 厦门大学 | A kind of motor behavior method for classifying modes, system and device |
CN109460812A (en) * | 2017-09-06 | 2019-03-12 | 富士通株式会社 | Average information analytical equipment, the optimization device, feature visualization device of neural network |
CN109726662A (en) * | 2018-12-24 | 2019-05-07 | 南京师范大学 | Multi-class human posture recognition method based on convolution sum circulation combination neural net |
CN109801200A (en) * | 2018-12-03 | 2019-05-24 | 国政通科技有限公司 | A kind of method and system of hierarchical detection |
CN109886109A (en) * | 2019-01-16 | 2019-06-14 | 南京邮电大学 | A kind of Activity recognition method based on deep learning |
CN110664412A (en) * | 2019-09-19 | 2020-01-10 | 天津师范大学 | Human activity recognition method facing wearable sensor |
CN111742327A (en) * | 2018-02-19 | 2020-10-02 | 博朗有限公司 | Apparatus and method for implementing positioning of a movable processing device |
CN112101235A (en) * | 2020-09-16 | 2020-12-18 | 济南大学 | Old people behavior identification and detection method based on old people behavior characteristics |
CN112434669A (en) * | 2020-12-14 | 2021-03-02 | 武汉纺织大学 | Multi-information fusion human behavior detection method and system |
CN112633467A (en) * | 2020-11-25 | 2021-04-09 | 超越科技股份有限公司 | Human behavior recognition method based on cat eye connection improved LSTM model |
CN112926553A (en) * | 2021-04-25 | 2021-06-08 | 北京芯盾时代科技有限公司 | Training method and device for motion detection network |
CN112932469A (en) * | 2021-01-26 | 2021-06-11 | 山西三友和智慧信息技术股份有限公司 | CNN + Transformer-based triaxial acceleration activity identification method |
CN113269400A (en) * | 2021-04-25 | 2021-08-17 | 贵州电网有限责任公司 | Low-voltage distribution network equipment state evaluation method based on historical operation and maintenance information |
CN114237394A (en) * | 2021-12-13 | 2022-03-25 | 广东乐心医疗电子股份有限公司 | Motion recognition method, device, equipment and medium |
GB2609542A (en) * | 2021-06-02 | 2023-02-08 | Nvidia Corp | Techniques for classification with neural networks |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105678222A (en) * | 2015-12-29 | 2016-06-15 | 浙江大学 | Human behavior identification method based on mobile equipment |
WO2016106383A2 (en) * | 2014-12-22 | 2016-06-30 | Robert Bosch Gmbh | First-person camera based visual context aware system |
CN105844239A (en) * | 2016-03-23 | 2016-08-10 | 北京邮电大学 | Method for detecting riot and terror videos based on CNN and LSTM |
CN106022239A (en) * | 2016-05-13 | 2016-10-12 | 电子科技大学 | Multi-target tracking method based on recurrent neural network |
-
2017
- 2017-04-28 CN CN201710292408.4A patent/CN107092894A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016106383A2 (en) * | 2014-12-22 | 2016-06-30 | Robert Bosch Gmbh | First-person camera based visual context aware system |
CN105678222A (en) * | 2015-12-29 | 2016-06-15 | 浙江大学 | Human behavior identification method based on mobile equipment |
CN105844239A (en) * | 2016-03-23 | 2016-08-10 | 北京邮电大学 | Method for detecting riot and terror videos based on CNN and LSTM |
CN106022239A (en) * | 2016-05-13 | 2016-10-12 | 电子科技大学 | Multi-target tracking method based on recurrent neural network |
Non-Patent Citations (1)
Title |
---|
MARCUS EDEL ET AL: "《An Advanced Method for Pedestrian Dead Reckoning using BLSTM-RNNs》", 《2015 INTERNATIONAL CONFERENCE ON INDOOR POSITIONING AND INDOOR NAVIGATION (IPIN)》 * |
Cited By (31)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109460812A (en) * | 2017-09-06 | 2019-03-12 | 富士通株式会社 | Average information analytical equipment, the optimization device, feature visualization device of neural network |
CN111771209A (en) * | 2018-02-19 | 2020-10-13 | 博朗有限公司 | Apparatus and method for classifying motion of a movable processing device |
CN111742327A (en) * | 2018-02-19 | 2020-10-02 | 博朗有限公司 | Apparatus and method for implementing positioning of a movable processing device |
CN111771208A (en) * | 2018-02-19 | 2020-10-13 | 博朗有限公司 | Apparatus and method for implementing positioning of a movable processing device |
CN108926338B (en) * | 2018-05-31 | 2019-06-18 | 中南民族大学 | Heart rate prediction technique and device based on deep learning |
CN108926338A (en) * | 2018-05-31 | 2018-12-04 | 中南民族大学 | Heart rate prediction technique and device based on deep learning |
CN109214285A (en) * | 2018-08-01 | 2019-01-15 | 浙江深眸科技有限公司 | Detection method is fallen down based on depth convolutional neural networks and shot and long term memory network |
CN109344960A (en) * | 2018-09-01 | 2019-02-15 | 哈尔滨工程大学 | A kind of DGRU neural network and its prediction model method for building up preventing data information loss |
CN109447162A (en) * | 2018-11-01 | 2019-03-08 | 山东大学 | A kind of real-time Activity recognition system and its working method based on Lora and Capsule |
CN109447162B (en) * | 2018-11-01 | 2021-09-24 | 山东大学 | Real-time behavior recognition system based on Lora and Capsule and working method thereof |
CN109447164B (en) * | 2018-11-01 | 2019-07-19 | 厦门大学 | A kind of motor behavior method for classifying modes, system and device |
US11551479B2 (en) | 2018-11-01 | 2023-01-10 | Xiamen University | Motion behavior pattern classification method, system and device |
WO2020088491A1 (en) * | 2018-11-01 | 2020-05-07 | 厦门大学 | Method, system, and device for classifying motion behavior mode |
CN109447164A (en) * | 2018-11-01 | 2019-03-08 | 厦门大学 | A kind of motor behavior method for classifying modes, system and device |
CN109801200A (en) * | 2018-12-03 | 2019-05-24 | 国政通科技有限公司 | A kind of method and system of hierarchical detection |
CN109726662A (en) * | 2018-12-24 | 2019-05-07 | 南京师范大学 | Multi-class human posture recognition method based on convolution sum circulation combination neural net |
CN109886109A (en) * | 2019-01-16 | 2019-06-14 | 南京邮电大学 | A kind of Activity recognition method based on deep learning |
CN109886109B (en) * | 2019-01-16 | 2022-08-30 | 南京邮电大学 | Behavior identification method based on deep learning |
CN110664412A (en) * | 2019-09-19 | 2020-01-10 | 天津师范大学 | Human activity recognition method facing wearable sensor |
CN112101235A (en) * | 2020-09-16 | 2020-12-18 | 济南大学 | Old people behavior identification and detection method based on old people behavior characteristics |
CN112633467A (en) * | 2020-11-25 | 2021-04-09 | 超越科技股份有限公司 | Human behavior recognition method based on cat eye connection improved LSTM model |
CN112434669B (en) * | 2020-12-14 | 2023-09-26 | 武汉纺织大学 | Human body behavior detection method and system based on multi-information fusion |
CN112434669A (en) * | 2020-12-14 | 2021-03-02 | 武汉纺织大学 | Multi-information fusion human behavior detection method and system |
CN112932469A (en) * | 2021-01-26 | 2021-06-11 | 山西三友和智慧信息技术股份有限公司 | CNN + Transformer-based triaxial acceleration activity identification method |
CN113269400B (en) * | 2021-04-25 | 2022-12-06 | 贵州电网有限责任公司 | Low-voltage distribution network equipment state evaluation method based on historical operation and maintenance information |
CN113269400A (en) * | 2021-04-25 | 2021-08-17 | 贵州电网有限责任公司 | Low-voltage distribution network equipment state evaluation method based on historical operation and maintenance information |
CN112926553B (en) * | 2021-04-25 | 2021-08-13 | 北京芯盾时代科技有限公司 | Training method and device for motion detection network |
CN112926553A (en) * | 2021-04-25 | 2021-06-08 | 北京芯盾时代科技有限公司 | Training method and device for motion detection network |
GB2609542A (en) * | 2021-06-02 | 2023-02-08 | Nvidia Corp | Techniques for classification with neural networks |
GB2609542B (en) * | 2021-06-02 | 2023-12-13 | Nvidia Corp | Techniques for classification with neural networks |
CN114237394A (en) * | 2021-12-13 | 2022-03-25 | 广东乐心医疗电子股份有限公司 | Motion recognition method, device, equipment and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107092894A (en) | A kind of motor behavior recognition methods based on LSTM models | |
Liu et al. | Skeleton-based action recognition using spatio-temporal LSTM network with trust gates | |
CN108764059B (en) | Human behavior recognition method and system based on neural network | |
CN106570477B (en) | Vehicle cab recognition model building method and model recognizing method based on deep learning | |
CN101807245B (en) | Artificial neural network-based multi-source gait feature extraction and identification method | |
CN103268495B (en) | Human body behavior modeling recognition methods based on priori knowledge cluster in computer system | |
CN103605972B (en) | Non-restricted environment face verification method based on block depth neural network | |
CN108764207A (en) | A kind of facial expression recognizing method based on multitask convolutional neural networks | |
CN109101938B (en) | Multi-label age estimation method based on convolutional neural network | |
CN106951867A (en) | Face identification method, device, system and equipment based on convolutional neural networks | |
CN108245172B (en) | Human body posture recognition method free of position constraint | |
CN109101876A (en) | Human bodys' response method based on long memory network in short-term | |
CN109815826A (en) | The generation method and device of face character model | |
CN110348364B (en) | Basketball video group behavior identification method combining unsupervised clustering and time-space domain depth network | |
CN105447473A (en) | PCANet-CNN-based arbitrary attitude facial expression recognition method | |
CN106778664A (en) | The dividing method and its device of iris region in a kind of iris image | |
CN110610158A (en) | Human body posture identification method and system based on convolution and gated cyclic neural network | |
CN110222634A (en) | A kind of human posture recognition method based on convolutional neural networks | |
CN109583331B (en) | Deep learning-based accurate positioning method for positions of wrist vein and mouth of person | |
CN106909938A (en) | Viewing angle independence Activity recognition method based on deep learning network | |
CN109376613A (en) | Video brainpower watch and control system based on big data and depth learning technology | |
CN110097029B (en) | Identity authentication method based on high way network multi-view gait recognition | |
CN110478883A (en) | A kind of body-building movement teaching and correction system and method | |
CN108073851A (en) | A kind of method, apparatus and electronic equipment for capturing gesture identification | |
CN104537273B (en) | A kind of drowned pattern intelligent inference system and method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170825 |