CN112183586A - Human body posture radio frequency identification method for on-line multi-task learning - Google Patents
Human body posture radio frequency identification method for on-line multi-task learning Download PDFInfo
- Publication number
- CN112183586A CN112183586A CN202010945467.9A CN202010945467A CN112183586A CN 112183586 A CN112183586 A CN 112183586A CN 202010945467 A CN202010945467 A CN 202010945467A CN 112183586 A CN112183586 A CN 112183586A
- Authority
- CN
- China
- Prior art keywords
- classifier
- posture
- task
- human body
- data
- 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.)
- Granted
Links
- 230000013016 learning Effects 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 19
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 10
- 238000011156 evaluation Methods 0.000 claims abstract description 7
- 230000036544 posture Effects 0.000 claims description 46
- 238000000819 phase cycle Methods 0.000 claims description 21
- 230000010363 phase shift Effects 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 10
- 230000000694 effects Effects 0.000 claims description 4
- 238000003491 array Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008878 coupling Effects 0.000 claims description 3
- 238000010168 coupling process Methods 0.000 claims description 3
- 238000005859 coupling reaction Methods 0.000 claims description 3
- 238000002790 cross-validation Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000005855 radiation Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000010801 machine learning Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000035045 associative learning Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K17/00—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations
- G06K17/0022—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations arrangements or provisious for transferring data to distant stations, e.g. from a sensing device
- G06K17/0029—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations arrangements or provisious for transferring data to distant stations, e.g. from a sensing device the arrangement being specially adapted for wireless interrogation of grouped or bundled articles tagged with wireless record carriers
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Computational Linguistics (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Computer Networks & Wireless Communication (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Psychiatry (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
A human body posture radio frequency identification method for on-line multi-task learning comprises the following steps: step 1: deploying a passive RFID electronic tag, an antenna and a reader-writer; step 2: collecting data, and training a standing posture classifier and a related task classifier; and step 3: and carrying out attitude recognition on the online target, and designing an evaluation index on a recognition result. The human body posture is recognized by using the RFID technology, a camera is not needed for collecting the portrait, the privacy is not influenced, various sensors are not needed to be worn, and the method is more convenient; the convolutional neural network is trained by using the multi-task learning method of online learning characteristics, and the classifier can be updated online in real time, so that the model generalization capability is stronger, and the recognition accuracy is higher.
Description
Technical Field
The invention relates to a human body posture radio frequency identification method for on-line multitask learning, in particular to a human body posture identification method for on-line multitask learning convolutional neural network based on a radio frequency technology.
Background
The gesture analysis of the human body has great significance for improving the living habits and the health of people. Many researches and developments have been made on human body posture recognition methods using Kinect, RGB cameras, wearable devices, or smart pads. The camera-based method has higher accuracy due to higher resolution, but it often causes people to worry about privacy issues. Wearable devices may be uncomfortable or require long-term maintenance, and the products need to be continuously optimized if the wearable devices are used for a long time.
RFID (Radio Frequency Identification) is a contactless automatic Identification technology. The radio frequency signal is used to identify the target without human intervention. The RFID is used for realizing low cost of human posture recognition, a camera and other sensors are not needed, but the machine learning method widely used in the existing RFID-based posture recognition method is the traditional methods such as decision tree, Bayes, K-neighborhood, random forest and ensemble learning, data cannot have missing values in the process of training data, but the actual situation may have the situation of data loss.
In the off-line learning of machine learning, training data in a data set is certainly available during model training, the model can be used for identifying things after training is finished, the on-line learning does not need to directly provide an available complete training set, and the on-line learning focuses on receiving real-time data and continuously updating a classifier. In order to improve the learning effect and promote the generalization ability, the invention adopts an online multi-task learning strategy based on deep learning, and the multi-task learning refers to a machine learning method which focuses more on learning a plurality of related tasks together based on shared representation. By performing associative learning on the relevant posture recognition tasks, information sharing at lower layers can prevent overfitting of the network.
Disclosure of Invention
In order to improve the above-mentioned shortcomings of the prior art, the present invention provides a human body posture radio frequency identification method for on-line multitask learning.
The human body posture recognition is realized by deploying the RFID tags indoors and combining an online multi-task deep convolutional network, a standing posture recognition is taken as a main task, a sitting posture recognition and a sleeping posture recognition are taken as related tasks, and a convolutional neural network model is trained.
A human body posture radio frequency identification method for on-line multi-task learning comprises the following steps:
step 1: deploying a passive RFID electronic tag, an antenna and a reader-writer;
step 2: collecting data, and training a standing posture classifier and a related task classifier;
and step 3: and carrying out attitude recognition on the online target, and designing an evaluation index on a recognition result.
In the step 1, the tag can be deployed on an indoor wall surface, a mattress or a key part of a human body according to which identification task is actually to be solved, and the reader-writer and the antenna are placed in an environment. The present invention places the antenna right against the tag. In order to prevent the mutual coupling from affecting the reception phase sequence by changing the radiation pattern of the tag antenna, the distance between adjacent tags should be larger than 5 cm.
The step 2 comprises the following steps:
2.1, the tester stands, sits and lies in different postures, records data of a period of time for different postures, and can obtain different phase sequence arrays of standing, sitting and lying, wherein each element of the array is the RSSI value of each time slot. When the tester changes posture, the corresponding RSSI value also changes, and the array data at each moment can be represented as a frame.
2.2, preprocessing data;
2.2.1, eliminating the constant phase shift caused by hardware in the RFID system. The raw set of phase sequence numbers is input, two adjacent phase measurements are compared, and a threshold is applied to detect and calibrate the phase shift, the threshold being set to coincide with the constant phase shift caused by the hardware. If the phase difference between two adjacent measurements exceeds a threshold, the phase shift is calibrated by adding or subtracting a threshold from the outlier.
2.2.2, eliminating the multipath effect in the indoor environment. And removing high-frequency noise in the measurement phase sequence by adopting a wavelet de-noising filter.
2.3, training a multi-task convolutional neural network based on the data in the step 2.2, performing associated training by taking a training standing posture classifier as a main task and taking a sitting posture and sleeping posture identification task as a related task, and updating the classifier on line by utilizing the correlation among the tasks.
The step 3 comprises the following steps:
3.1, taking the data collected at each moment as current frame data, and extracting characteristic parameters
Dividing the collected phase sequence into a plurality of segments by adopting a sliding window, extracting the characteristics of 'phase sequence average value', 'phase sequence median', 'signal intensity S' and 'information quantity Q' of each segment, wherein the calculation modes of S and Q are respectively shown as a formula 1 and a formula 2, wherein alpha is a fast Fourier transform coefficient, N is the size of the sliding window, N is the size of the sliding window, and N is the size of the sliding windowiIs a fast fourier transform coefficient normalization value.
Equation 1:
equation 2:
and 3.2, inputting the extracted characteristic parameters into the classifier obtained in the step 2, performing online learning on the characteristics to obtain a current posture recognition result, and updating the classifier online at the same time, so that the recognition result of the multitask convolutional neural network model is more accurate.
3.3, classifier performance was evaluated using cross-validation and using four metrics of accuracy, F1 score, accuracy and recall. In evaluating the accuracy of feature segmentation for each pose, the amount of deviation of each boundary from the theoretical value is used as an evaluation index, with the larger the amount of deviation, the larger the error.
The invention has the advantages that:
1) the human body posture is recognized by using the RFID technology, a camera is not needed for collecting the portrait, the privacy is not influenced, various sensors are not needed to be worn, and the method is more convenient;
2) the convolutional neural network is trained by using a multi-task learning method of online learning characteristics, so that the classifier can be updated online in real time, the generalization capability of the model is stronger, and the recognition accuracy is higher;
description of the drawings:
FIG. 1 is a basic flow diagram of the process of the present invention;
FIG. 2 is a flow chart of the present invention for data acquisition and classifier training;
FIG. 3 is a flow chart of an online gesture recognition method of the present invention.
Detailed Description
The invention provides a human body posture radio frequency identification method for on-line multitask learning, which realizes human body posture identification by deploying RFID tags indoors and combining an on-line multitask deep convolution network mode, and trains a convolution neural network model by taking standing posture identification as a main task and sitting posture identification and sleeping posture identification as related tasks.
A human body posture radio frequency identification method for on-line multitask learning is shown as a basic flow chart of the method in the invention as shown in figure 1, and comprises the following steps:
step 1: deploying a passive RFID electronic tag, an antenna and a reader-writer;
step 2: collecting data, and training a standing posture classifier and a related task classifier;
and step 3: and carrying out attitude recognition on the online target, and designing an evaluation index on a recognition result.
In the step 1, the tag can be deployed on an indoor wall surface, a mattress or a key part of a human body according to which identification task is actually to be solved, and the reader-writer and the antenna are placed in an environment. The embodiment deploys the tags on the wall and floor with the antenna placed against the tag. In order to prevent the mutual coupling from influencing the receiving phase sequence due to the change of the radiation mode of the tag antenna, the distance between the adjacent tags is more than 5 cm, and the distance between the adjacent tags is set to be 10 cm by the invention.
The flowchart of step 2 is shown in fig. 2, and includes the following steps:
2.1, the tester stands, sits and lies in different postures, records data of a period of time for different postures, and can obtain different phase sequence arrays of standing, sitting and lying, wherein each element of the array is the RSSI value of each time slot. When the tester changes posture, the corresponding RSSI value also changes, and the array data at each moment can be represented as a frame.
2.2 data preprocessing
2.2.1, eliminating the constant phase shift caused by hardware in the RFID system. The raw set of phase sequence numbers is input, two adjacent phase measurements are compared, and a threshold is applied to detect and calibrate the phase shift, the threshold being set to coincide with the constant phase shift caused by the hardware. If the phase difference between two adjacent measurements exceeds a threshold, the phase shift is calibrated by adding or subtracting a threshold from the outlier.
2.2.2, eliminating the multipath effect in the indoor environment. And removing high-frequency noise in the measurement phase sequence by adopting a wavelet de-noising filter.
2.3, training a multi-task convolutional neural network based on the data in the step 2.2, performing associated training by taking a training standing posture classifier as a main task and taking a sitting posture and sleeping posture identification task as a related task, and updating the classifier on line by utilizing the correlation among the tasks.
The step 3 is shown in fig. 3, and comprises the following steps:
3.1, taking the data collected at each moment as current frame data, and extracting characteristic parameters
Dividing the collected phase sequence into a plurality of segments by adopting a sliding window, extracting the characteristics of 'phase sequence average value', 'phase sequence median', 'signal intensity S' and 'information quantity Q' of each segment, wherein the calculation modes of S and Q are respectively shown as a formula 1 and a formula 2, wherein alpha is a fast Fourier transform coefficient, N is the size of the sliding window, N is the size of the sliding window, and N is the size of the sliding windowiIs a fast fourier transform coefficient normalization value.
Equation 1:
equation 2:
and 3.2, inputting the extracted characteristic parameters into the classifier obtained in the step 2, performing online learning on the characteristics to obtain a current posture recognition result, and updating the classifier online at the same time, so that the recognition result of the multitask convolutional neural network model is more accurate.
3.3, classifier performance was evaluated using cross-validation and using four metrics of accuracy, F1 score, accuracy and recall. In evaluating the accuracy of feature segmentation for each pose, the amount of deviation of each boundary from the theoretical value is used as an evaluation index, with the larger the amount of deviation, the larger the error.
Claims (2)
1. A human body posture radio frequency identification method for on-line multitask learning comprises the following steps:
step 1: deploying a passive RFID electronic tag, an antenna and a reader-writer;
step 2: collecting data, and training a standing posture classifier and a related task classifier; the method specifically comprises the following steps:
2.1, a tester stands, sits and lies in different postures, records data of a period of time for different postures, and can obtain different phase sequence arrays of standing, sitting and lying, wherein each element of the array is an RSSI value on each time slot; when the tester changes the posture, the corresponding RSSI value also changes, and the array data at each moment can be represented as a frame;
2.2, preprocessing data;
2.2.1, deleting the constant phase shift caused by hardware in the radio frequency identification system; inputting an original phase sequence array, comparing two adjacent phase measurement values, and applying a threshold value to detect and calibrate phase shift, wherein the threshold value is set to be consistent with constant phase shift caused by hardware; if the phase difference between two adjacent measurements exceeds a threshold, calibrating the phase shift by adding or subtracting a threshold from the outlier;
2.2.2, eliminating the multipath effect in the indoor environment; removing high-frequency noise in the measurement phase sequence by adopting a wavelet denoising filter;
2.3, training a multi-task convolutional neural network based on the data in the step 2.2, performing associated training by taking a training standing posture classifier as a main task and taking a sitting posture and sleeping posture identification task as a related task, and updating the classifier on line by utilizing the correlation among the tasks;
and step 3: carrying out attitude identification on the online target, and designing an evaluation index on an identification result;
3.1, taking the data collected at each moment as current frame data, and extracting characteristic parameters
Dividing the collected phase sequence into a plurality of segments by adopting a sliding window, extracting the characteristics of 'phase sequence average value', 'phase sequence median', 'signal intensity S' and 'information quantity Q' of each segment, wherein the calculation modes of S and Q are respectively shown as a formula 1 and a formula 2, wherein alpha is a fast Fourier transform coefficient, N is the size of the sliding window, N is the size of the sliding window, and N is the size of the sliding windowiIs a fast fourier transform coefficient normalization value;
equation 1:
equation 2:
3.2, inputting the extracted characteristic parameters into the classifier obtained in the step 2, performing online learning on the characteristics to obtain a current posture recognition result, and updating the classifier online at the same time, so that the recognition result of the multitask convolutional neural network model is more accurate;
3.3, evaluating the performance of the classifier by using cross validation and four measurement standards of accuracy, F1 score, precision and recall ratio; in evaluating the accuracy of feature segmentation for each pose, the amount of deviation of each boundary from the theoretical value is used as an evaluation index, with the larger the amount of deviation, the larger the error.
2. The human body posture radio frequency identification method for on-line multitask learning as claimed in claim 1, characterized in that: in the step 1, the tag is deployed on an indoor wall surface, a mattress or a key part of a human body according to which identification task is actually solved, and the reader-writer and the antenna are arranged in an environment; placing the antenna against the tag; in order to prevent the mutual coupling from affecting the reception phase sequence by changing the radiation pattern of the tag antenna, the distance between adjacent tags should be larger than 5 cm.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010945467.9A CN112183586B (en) | 2020-09-10 | 2020-09-10 | Human body posture radio frequency identification method for online multitask learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010945467.9A CN112183586B (en) | 2020-09-10 | 2020-09-10 | Human body posture radio frequency identification method for online multitask learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112183586A true CN112183586A (en) | 2021-01-05 |
CN112183586B CN112183586B (en) | 2024-04-02 |
Family
ID=73920466
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010945467.9A Active CN112183586B (en) | 2020-09-10 | 2020-09-10 | Human body posture radio frequency identification method for online multitask learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112183586B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113273998A (en) * | 2021-07-08 | 2021-08-20 | 南京大学 | Human body sleep information acquisition method and device based on RFID label matrix |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103576860A (en) * | 2013-10-30 | 2014-02-12 | 山东省射频识别应用工程技术研究中心有限公司 | Electronic tag identification method and device based on 3D (3-dimensional) motion postures |
CN107832647A (en) * | 2017-10-20 | 2018-03-23 | 南京邮电大学 | A kind of phase type gesture identification method based on passive RF label |
US20180189529A1 (en) * | 2016-05-13 | 2018-07-05 | Nanjing University | Book placing attitude recognition method based on rssi signal value of rfid |
CN108920993A (en) * | 2018-03-23 | 2018-11-30 | 武汉雷博合创电子科技有限公司 | A kind of pedestrian's gesture recognition method and system based on radar and multiple networks fusion |
-
2020
- 2020-09-10 CN CN202010945467.9A patent/CN112183586B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103576860A (en) * | 2013-10-30 | 2014-02-12 | 山东省射频识别应用工程技术研究中心有限公司 | Electronic tag identification method and device based on 3D (3-dimensional) motion postures |
US20180189529A1 (en) * | 2016-05-13 | 2018-07-05 | Nanjing University | Book placing attitude recognition method based on rssi signal value of rfid |
CN107832647A (en) * | 2017-10-20 | 2018-03-23 | 南京邮电大学 | A kind of phase type gesture identification method based on passive RF label |
CN108920993A (en) * | 2018-03-23 | 2018-11-30 | 武汉雷博合创电子科技有限公司 | A kind of pedestrian's gesture recognition method and system based on radar and multiple networks fusion |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113273998A (en) * | 2021-07-08 | 2021-08-20 | 南京大学 | Human body sleep information acquisition method and device based on RFID label matrix |
Also Published As
Publication number | Publication date |
---|---|
CN112183586B (en) | 2024-04-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11902857B2 (en) | Handling concept drift in Wi-Fi-based localization | |
Rakun et al. | Detecting fruits in natural scenes by using spatial-frequency based texture analysis and multiview geometry | |
CN112036433B (en) | CNN-based Wi-Move behavior sensing method | |
CN111505632B (en) | Ultra-wideband radar action attitude identification method based on power spectrum and Doppler characteristics | |
CN106559749B (en) | Multi-target passive positioning method based on radio frequency tomography | |
CN105718971B (en) | A kind of multiple target passive type indoor activity recognition methods based on RFID | |
KR101433472B1 (en) | Apparatus, method and computer readable recording medium for detecting, recognizing and tracking an object based on a situation recognition | |
CN108734283A (en) | Nerve network system | |
CN111368635A (en) | Millimeter wave-based multi-person gait recognition method and device | |
Wang et al. | Multi-classification of UWB signal propagation channels based on one-dimensional wavelet packet analysis and CNN | |
CN111586605A (en) | KNN indoor target positioning method based on adjacent weighted self-adaptive k value | |
CN112183586B (en) | Human body posture radio frequency identification method for online multitask learning | |
CN106778491B (en) | The acquisition methods and equipment of face 3D characteristic information | |
Selvakumar et al. | Automated mango leaf infection classification using weighted and deep features with optimized recurrent neural network concept | |
Zhong et al. | A climate adaptation device-free sensing approach for target recognition in foliage environments | |
Yang et al. | Plot-scale rice grain yield estimation using UAV-based remotely sensed images via CNN with time-invariant deep features decomposition | |
CN110569895B (en) | Micro Doppler radar human body action classification method of convolutional neural network | |
CN113456061A (en) | Sleep posture monitoring method and system based on wireless signals | |
Turetta et al. | Practical identity recognition using wifi's channel state information | |
Wu et al. | Device-free human activity recognition with identity-based transfer mechanism | |
CN113739790B (en) | Passive indoor positioning system and positioning method | |
CN115660042A (en) | Three-dimensional sleeping posture detection method based on RFID (radio frequency identification) equipment | |
CN112883905B (en) | Human behavior recognition method based on coarse-grained time-frequency features and multi-layer fusion learning | |
Prabhu et al. | A Deep Learning Approach to Identify Defects in Coffee Leaves Using Convoluional Neural Network | |
CN112086165A (en) | Upper limb rehabilitation monitoring method and system based on deep learning |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |