[go: up one dir, main page]

CN112347923A - A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network - Google Patents

A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network Download PDF

Info

Publication number
CN112347923A
CN112347923A CN202011229272.0A CN202011229272A CN112347923A CN 112347923 A CN112347923 A CN 112347923A CN 202011229272 A CN202011229272 A CN 202011229272A CN 112347923 A CN112347923 A CN 112347923A
Authority
CN
China
Prior art keywords
pedestrian
trajectory
latent variable
pedestrians
generator
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.)
Withdrawn
Application number
CN202011229272.0A
Other languages
Chinese (zh)
Inventor
杨彪
何才臻
徐黎明
闫国成
吕继东
陈阳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu China Israel Industrial Technology Research Institute
Changzhou University
Original Assignee
Jiangsu China Israel Industrial Technology Research Institute
Changzhou University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu China Israel Industrial Technology Research Institute, Changzhou University filed Critical Jiangsu China Israel Industrial Technology Research Institute
Priority to CN202011229272.0A priority Critical patent/CN112347923A/en
Publication of CN112347923A publication Critical patent/CN112347923A/en
Withdrawn legal-status Critical Current

Links

Images

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
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
    • 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
    • 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
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • 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
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Business, Economics & Management (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Biomedical Technology (AREA)
  • Health & Medical Sciences (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Multimedia (AREA)
  • Human Computer Interaction (AREA)
  • Development Economics (AREA)
  • Evolutionary Biology (AREA)
  • Game Theory and Decision Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention relates to a roadside end pedestrian track generation algorithm based on an confrontation generation network, which generates a multi-mode predicted track by utilizing a social attention mechanism and a pedestrian track latent variable; through the confrontation generation training of the track generator and the discriminator, the capabilities of the generator and the discriminator are continuously optimized, and the precision of the track generated by the generator is improved; the method comprises the steps that a social attention mechanism based on head orientation is provided, the head orientation of a pedestrian is obtained through the last speed direction of the pedestrian, a cosine value of an included angle between the pedestrians is calculated according to head orientation information, soft attention and hard attention mechanisms optimize the output of the social attention mechanism by using the calculated angle information, and the output is converged through a maximum pooling layer; a new latent variable generating method is proposed, two feedforward neural networks are used for learning latent variables from pedestrian historical tracks and observation tracks respectively, the input of the latent variable generator comprises position, speed and acceleration, and the distribution of three types of latent variables is generated from the three types of input respectively.

Description

Roadside end pedestrian track prediction algorithm based on confrontation generation network
Technical Field
The invention relates to the technical field of automatic driving, in particular to pedestrian trajectory prediction, and provides a roadside end pedestrian trajectory prediction algorithm based on a confrontation generation network.
Background
With the continuous development of the automatic navigation of the robot and the automatic driving technology of the automobile, the unmanned technology gets wide attention and has bright application prospect; the unmanned vehicle can bring convenience to the life of people, but the unmanned vehicle needs to monitor the motion trail of pedestrians on a road and predict the future motion trail of the pedestrians during driving, so that collision with the pedestrians is avoided; in order to better predict the motion trail of the pedestrian, the unmanned vehicle needs to process the observed pedestrian trail data, learn the rule of the motion of the pedestrian and predict the next motion state of the pedestrian according to the rule; the challenge of accurately predicting the pedestrian motion trajectory comes from the complexity of human behavior and its own intentions and variety of external stimuli; pedestrian motion behavior may be driven by its own target intent, the existence of action interactions between surrounding objects, social relationships, social rules and norms, or its topological, geometric and semantic environment, most of which are not directly visible, need to be inferred from complex laws of motion, or modeled from contextual information; how to let the unmanned vehicle learn the potential motion law is the key for accurately predicting the pedestrian track;
due to the fact that behaviors of pedestrians are random, whether the pedestrians are machines or humans, future tracks of the pedestrians cannot be predicted accurately; the pedestrian's trajectory is influenced by the surrounding environment, such as person-to-person, person-to-object, which is potentially undescribable; however, the future track of the pedestrian is always influenced by the motion of people and objects in front of the pedestrian, and the common knowledge is utilized to be beneficial to simulating the social interaction behavior of the pedestrian, so that the future motion track of the pedestrian is well predicted;
the motion modes of pedestrians are complex and diverse, the complex pedestrian motion is difficult to describe by a dynamic model, and a common method for modeling the general motion of a maneuvering target is to define and fuse different typical motion modes, each mode is described by different dynamic states; the patterns may be linear movements, turning maneuvers or sudden accelerations, forming over time a sequence capable of describing complex movement behaviour; the diversity of pedestrian motion patterns in pedestrian trajectory prediction must also be considered;
disclosure of Invention
The technical problem to be solved by the invention is as follows: in order to solve the problems that the motion modes of pedestrians are complex and various and the complex pedestrian motion is difficult to describe by a dynamic model in the prior art, a roadside end pedestrian track prediction algorithm based on a confrontation generation network is provided.
The technical scheme adopted by the invention for solving the technical problems is as follows: a roadside end pedestrian trajectory prediction algorithm based on a confrontation generation network comprises the following steps:
s10: encoding the input track using an encoder;
s20: calculating the social attention of the pedestrian by utilizing the head orientation of the pedestrian;
s30: applying a latent variable predictor to generate a predictable latent variable distribution;
s40: generating a predicted future trajectory of the pedestrian;
s50: optimizing the pedestrian trajectory generated by the generator using a discriminator;
the step S30 includes the following steps:
s31: designing a latent variable predictor;
s32: and predicting the potential variable distribution of the pedestrian by using a latent variable predictor.
Further, in step S31: the latent variable predictor consists of two feedforward neural networks defined as follows:
Figure BDA0002764619690000021
Figure BDA0002764619690000022
wherein Ψ (-) and
Figure BDA0002764619690000031
is a feed-forward neural network that is,
Figure BDA0002764619690000032
and
Figure BDA0002764619690000033
are the parameters of the two feedforward neural networks respectively,
Figure BDA0002764619690000034
and
Figure BDA0002764619690000035
is the k-th type input of the latent variable predictor.
Further, in step S32: k is 1, 2 and 3, and respectively represents the position, speed and acceleration of the pedestrian, the position reveals the layout of the potential scene, the speed reflects the motion mode of different pedestrians, and the acceleration shows the motion intensity of the pedestrian; the latent variable predictor estimates the latent distribution of three variables from the three inputs; gaussian random noise is used for generating multi-mode output, and finally, the three kinds of latent variable distribution and the Gaussian random noise are fused together to finally form latent variable distribution parameters in a training stage;
in the testing stage, a latent variable predictor predicts the latent variable distribution from the observation track of the pedestrian, the latent variable predictor inputs the position, speed and acceleration information of the pedestrian, can respectively predict the latent variable distribution of the position, speed and acceleration of the pedestrian from the three types of input, and combines the three types of latent variables and Gaussian random noise to form a final latent variable which is output to a track generator;
in the training process, the latent variable loss function is used for measuring the difference between the latent variable distribution of the observed track and the latent variable distribution of the real track, and KL divergence is used for calculating the error, wherein the formula is as follows:
Figure BDA0002764619690000036
wherein
Figure BDA0002764619690000037
And
Figure BDA0002764619690000038
respectively representing the latent variable distribution of the observed track and the latent variable distribution of the real track.
Further, the step 1 specifically includes the following steps:
s11: processing input track data: the input trace being a series of time-series trace points
Figure BDA0002764619690000039
Wherein
Figure BDA00027646196900000310
Is the position coordinate of the target i at time t; the position coordinates of each track at different moments are sent into a coding network;
s12: converting two-dimensional position information into multi-dimensional vector of fixed length by using single-layer multi-layer perceptron
Figure BDA00027646196900000311
The definition of the multi-layer perceptron is as follows:
Figure BDA00027646196900000312
where φ (-) is a multi-layered perceptron using a ReLU nonlinear activation function, WeeIs a parameter of the multi-layer perceptron;
s13: sending the multidimensional vector into a coder based on a long-term and short-term memory network to generate a hidden state of the pedestrian movement
Figure BDA0002764619690000041
The encoder long short term memory network (LSTM) is defined as follows:
Figure BDA0002764619690000042
where LSTM (. beta.) is a long-short term memory network, WencoderThe parameter is a parameter of a long-term and short-term memory network of the encoder, and the parameter can be shared among all pedestrians in the same scene.
Further, the step 2 specifically includes the following steps:
s21: calculating the azimuth angle between the pedestrians: taking the speed of the last position of the pedestrian as the future speed of the pedestrian, taking the direction of the speed of the last position of the pedestrian as the head direction and the track motion direction, and calculating the cosine value of the azimuth angle between the pedestrians by using the head directions of all the pedestrians as follows:
Figure BDA0002764619690000043
where n is the number of all pedestrians in the same scene, bijRepresenting the included angle between the pedestrian i and the pedestrian j;
s22: designing an attention mechanism: designing a soft attention mechanism and a hard attention mechanism according to the cosine values of the azimuth included angles among the pedestrians; the effect of one pedestrian on another decreases as the azimuthal cosine value between them increases; the hard attention mechanism uses a matrix H with the same shape as cos (beta)AIs represented by HAEach of the elements hijAre all set to 0 or 1, when the row isWhen the cosine value of the azimuth included angle between the people is greater than the preset threshold value 0.2, the corresponding attention weight hij1, when the cosine value of the azimuth included angle between the pedestrians is less than the preset threshold value 0.2, the corresponding attention weight hijIs 0; the soft attention mechanism and the hard attention mechanism calculate attention weights through thresholds; adaptive computation of correlations between pedestrians for a soft attention mechanism, weight S for the soft attention mechanismAThe calculation formula of (a) is as follows:
Figure BDA0002764619690000044
where δ (-) denotes a sigmoid activation function,
Figure BDA0002764619690000051
represents 1 × 1 convolutional layers;
the soft and hard attention machine is used for the output of the second multilayer perceptron, the soft attention machine and the hard attention machine are used for optimizing the output of the second multilayer perceptron, and the attention machine is converged through the largest pooling layer to obtain the output
Figure BDA0002764619690000052
Further, the step 4 specifically includes the following steps:
s41: output of social attention module
Figure BDA0002764619690000053
And the output of latent variable predictor
Figure BDA0002764619690000054
Hidden from pedestrian movement
Figure BDA0002764619690000055
Make a splice
S42: the splicing result is input into a decoder based on a long-term and short-term memory network to obtain a new track hidden state fused with various information
Figure BDA0002764619690000056
The long-short term memory network of the decoder is defined as follows:
Figure BDA0002764619690000057
where LSTM (. beta.) is a long-short term memory network, WdecoderThe parameters of the decoder long-term and short-term memory network can be shared among all pedestrians in the same scene;
s43: decoding the new hidden state by using a multilayer perceptron to obtain future track coordinates of the pedestrian: the multi-layer perceptron is defined as follows:
Figure BDA0002764619690000058
where γ (-) is a multi-layered perceptron using the ReLU nonlinear activation function,
Figure BDA0002764619690000059
is the future position coordinate of the pedestrian, and the output prediction result is a series of position coordinates
Figure BDA00027646196900000510
Wherein T isobsIs the length of the predicted trajectory; the invention adopts multi-mode output, the track generator outputs m tracks at a time, and 2 norm loss functions are used for calculating the deviation between the m tracks and the true value, and the expression is as follows:
Figure BDA00027646196900000511
wherein
Figure BDA00027646196900000512
Is the real track of the pedestrian,
Figure BDA00027646196900000513
is the m-th generationThe predicted future trajectory of the pedestrian is set to m-20 in the present invention.
Further, the step 5 specifically includes the following steps:
s51: inputting the trajectory generated by the generator and the real trajectory of the pedestrian into the discriminator
S52: the discriminator discriminates whether the input trajectory is a trajectory generated by the generator or a real trajectory: the discriminator uses an encoder based on a long-short term memory network to encode a real track and a generated track, a multi-layer perceptron is applied to a hidden state output by the encoder to obtain a classification score, under an ideal condition, the discriminator learns social rules of the pedestrian track, and the track which does not accord with the rules is judged to be false by the discriminator;
the penalty function against generative training is expressed as follows:
Figure BDA0002764619690000061
where D is the discriminator, G is the generator, z is the latent variable distribution parameter, x is the trajectory data of the data,
Figure BDA0002764619690000062
is the kth input (position, velocity, acceleration) of the observed trajectory latent variable predictor; through game training of the generator and the discriminator, the generator can finally generate samples which are similar to a training set and accord with social rules; because the generator learns a probability distribution similar to that of the training set, each sampling can give different reasonable samples, and therefore the probability distribution can be used for predicting multiple possibilities;
the total loss function is composed of three parts, wherein one part is a training loss function generated by confrontation, one part is the KL divergence of latent variable distribution, and the other part is the deviation between a predicted value and a true value; the total loss function weight is defined as follows:
Figure BDA0002764619690000063
wherein alpha and beta are respectively set as numbers between 1 and 10, and specific values are obtained by cross validation on a reference data set; during training, a generator and a discriminator are iteratively trained, the batch processing size is set to be 64, 600 epochs, the learning rate is set to be 0.001, and an Adam optimizer is used for optimizing parameters.
The invention has the beneficial effects that the roadside end pedestrian track prediction algorithm based on the confrontation generation network (1) provides a social attention module, the module utilizes the correlation between the head orientation and the track prediction, and the attention mechanism improves the social interaction capturing capability of the social pooling layer under different scenes;
(2) a new latent variable predictor is provided, which can estimate the latent variable with rich knowledge to better predict the track; only the input of the prediction variable is extracted from the trajectory data, thus only little computational overhead is added;
(3) embedding the social attention focusing module and the latent variable predictive variable into an confrontation generating network framework to generate multi-mode output acceptable by social rules;
drawings
The invention is further illustrated with reference to the following figures and examples.
FIG. 1 is a schematic diagram of a challenge generation training strategy proposed in the present invention;
FIG. 2 is a schematic diagram of a generator proposed in the present invention
FIG. 3 is a schematic representation of latent variable prediction proposed in the present invention;
FIG. 4 is a schematic diagram of the discriminator proposed in the present invention;
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic views illustrating only the basic structure of the present invention in a schematic manner, and thus show only the constitution related to the present invention.
1-4, the roadside end pedestrian trajectory prediction algorithm based on the confrontation generation network includes the following steps:
s10: encoding the input track using an encoder;
s20: calculating the social attention of the pedestrian by utilizing the head orientation of the pedestrian;
s30: applying a latent variable predictor to generate a predictable latent variable distribution;
s40: generating a predicted future trajectory of the pedestrian;
s50: optimizing the pedestrian trajectory generated by the generator using a discriminator;
the step 1 specifically comprises the following steps:
s11: processing input track data: the input trace being a series of time-series trace points
Figure BDA0002764619690000081
Wherein
Figure BDA0002764619690000082
Is the position coordinate of the target i at time t; the position coordinates of each track at different moments are sent into a coding network;
s12: converting two-dimensional position information into multi-dimensional vector of fixed length by using single-layer multi-layer perceptron
Figure BDA0002764619690000083
The definition of the multi-layer perceptron is as follows:
Figure BDA0002764619690000084
where φ (-) is a multi-layered perceptron using a ReLU nonlinear activation function, WeeIs a parameter of the multi-layer perceptron;
s13: sending the multidimensional vector into a coder based on a long-term and short-term memory network to generate a hidden state of the pedestrian movement
Figure BDA0002764619690000085
The encoder long short term memory network (LSTM) is defined as follows:
Figure BDA0002764619690000086
where LSTM (. beta.) is a long-short term memory network, WencoderThe parameter is a parameter of a long-term and short-term memory network of the encoder, and the parameter can be shared among all pedestrians in the same scene.
Further, the step 2 specifically includes the following steps:
s21: calculating the azimuth angle between the pedestrians: the invention is based on the fact that the future trajectory of a pedestrian is always influenced by the front crowd and not by the rear crowd; taking the speed of the last position of the pedestrian as the future speed of the pedestrian, taking the direction of the speed of the last position of the pedestrian as the head direction and the track motion direction, and calculating the cosine value of the azimuth angle between the pedestrians by using the head directions of all the pedestrians as follows:
Figure BDA0002764619690000087
where n is the number of all pedestrians in the same scene, bijRepresenting the included angle between the pedestrian i and the pedestrian j;
s22: designing an attention mechanism: designing a soft attention mechanism and a hard attention mechanism according to the cosine values of the azimuth included angles among the pedestrians; the effect of one pedestrian on another decreases as the azimuthal cosine value between them increases; the hard attention mechanism uses a matrix H with the same shape as cos (beta)AIs represented by HAEach of the elements hijIs set to be 0 or 1, when the cosine value of the azimuth angle between the pedestrians is greater than the preset threshold value 0.2, the corresponding attention weight hij1, when the cosine value of the azimuth included angle between the pedestrians is less than the preset threshold value 0.2, the corresponding attention weight hijIs 0; the soft attention mechanism and the hard attention mechanism calculate attention weights through thresholds; adaptive computation of correlations between pedestrians for a soft attention mechanism, weight S for the soft attention mechanismAThe calculation formula of (a) is as follows:
Figure BDA0002764619690000091
where δ (-) denotes a sigmoid activation function,
Figure BDA0002764619690000092
represents 1 × 1 convolutional layers;
the soft and hard attention machine is used for the output of the second multilayer perceptron, the soft attention machine and the hard attention machine are used for optimizing the output of the second multilayer perceptron, and the attention machine is converged through the largest pooling layer to obtain the output
Figure BDA0002764619690000093
The step S30 includes the following steps:
s31: designing a latent variable predictor;
s32: and predicting the potential variable distribution of the pedestrian by using a latent variable predictor.
In step S31: the invention applies a latent variable predictor to generate a predictable latent variable distribution, which is a method for predicting latent variable distribution parameters in a data-driven manner; potential variable distribution parameters can be predicted from the observation track and the real track of the pedestrian in a training stage by a potential variable generator, so that a potential motion rule can be learned; the latent variable predictor consists of two feedforward neural networks defined as follows:
Figure BDA0002764619690000094
Figure BDA0002764619690000095
wherein Ψ (-) and
Figure BDA0002764619690000096
is a feed-forward neural network that is,
Figure BDA0002764619690000097
and
Figure BDA0002764619690000098
are the parameters of the two feedforward neural networks respectively,
Figure BDA0002764619690000099
and
Figure BDA00027646196900000910
is the k-th type input of the latent variable predictor.
Further, in step S32: k is 1, 2 and 3, and respectively represents the position, speed and acceleration of the pedestrian, the position reveals the layout of the potential scene, the speed reflects the motion mode of different pedestrians, and the acceleration shows the motion intensity of the pedestrian; the latent variable predictor estimates the latent distribution of three variables from the three inputs; gaussian random noise is used for generating multi-mode output, and finally, the three kinds of latent variable distribution and the Gaussian random noise are fused together to finally form latent variable distribution parameters in a training stage;
in the testing stage, a latent variable predictor predicts the latent variable distribution from the observation track of the pedestrian, the latent variable predictor inputs the position, speed and acceleration information of the pedestrian, can respectively predict the latent variable distribution of the position, speed and acceleration of the pedestrian from the three types of input, and combines the three types of latent variables and Gaussian random noise to form a final latent variable which is output to a track generator;
in the training process, the latent variable loss function is used for measuring the difference between the latent variable distribution of the observed track and the latent variable distribution of the real track, and KL divergence is used for calculating the error, wherein the formula is as follows:
Figure BDA0002764619690000101
wherein
Figure BDA0002764619690000102
And
Figure BDA0002764619690000103
respectively representing the latent variable distribution of the observed track and the latent variable distribution of the real track.
Further, the step 4 specifically includes the following steps:
s41: the interaction between pedestrians is obtained by the social attention module, the pedestrian motion latent variable distribution is obtained by the latent variable predictor, and the output of the social attention module is output
Figure BDA0002764619690000104
And the output of latent variable predictor
Figure BDA0002764619690000105
Hidden from pedestrian movement
Figure BDA0002764619690000106
Make a splice
S42: the splicing result is input into a decoder based on a long-term and short-term memory network to obtain a new track hidden state fused with various information
Figure BDA0002764619690000107
The long-short term memory network of the decoder is defined as follows:
Figure BDA0002764619690000108
where LSTM (. beta.) is a long-short term memory network, WdecoderThe parameters of the decoder long-term and short-term memory network can be shared among all pedestrians in the same scene;
s43: decoding the new hidden state by using a multilayer perceptron to obtain future track coordinates of the pedestrian: the multi-layer perceptron is defined as follows:
Figure BDA0002764619690000111
where γ (-) is a multi-layered perceptron using the ReLU nonlinear activation function,
Figure BDA0002764619690000112
is the future position coordinate of the pedestrian, and the output prediction result is a series of position coordinates
Figure BDA0002764619690000113
Wherein T isobsIs the length of the predicted trajectory; the invention adopts multi-mode output, the track generator outputs m tracks at a time, and 2 norm loss functions are used for calculating the deviation between the m tracks and the true value, and the expression is as follows:
Figure BDA0002764619690000114
wherein
Figure BDA0002764619690000115
Is the real track of the pedestrian,
Figure BDA0002764619690000116
is the predicted future trajectory of the pedestrian by the mth generator, and m is set to 20 in the present invention.
Further, the step 5 specifically includes the following steps:
s51: inputting the trajectory generated by the generator and the real trajectory of the pedestrian into the discriminator
S52: the discriminator discriminates whether the input trajectory is a trajectory generated by the generator or a real trajectory: the discriminator uses an encoder based on a long-short term memory network to encode a real track and a generated track, a multi-layer perceptron is applied to a hidden state output by the encoder to obtain a classification score, under an ideal condition, the discriminator learns social rules of the pedestrian track, and the track which does not accord with the rules is judged to be false by the discriminator;
the penalty function against generative training is expressed as follows:
Figure BDA0002764619690000117
where D is the discriminator, G is the generator, z is the latent variable distribution parameter, x is the trajectory data of the data,
Figure BDA0002764619690000118
is the kth input (position, velocity, acceleration) of the observed trajectory latent variable predictor; through game training of the generator and the discriminator, the generator can finally generate samples which are similar to a training set and accord with social rules; because the generator learns a probability distribution similar to that of the training set, each sampling can give different reasonable samples, and therefore the probability distribution can be used for predicting multiple possibilities;
the total loss function is composed of three parts, wherein one part is a training loss function generated by confrontation, one part is the KL divergence of latent variable distribution, and the other part is the deviation between a predicted value and a true value; the total loss function weight is defined as follows:
Figure BDA0002764619690000121
wherein alpha and beta are respectively set as numbers between 1 and 10, and specific values are obtained by cross validation on a reference data set; during training, a generator and a discriminator are iteratively trained, the batch processing size is set to be 64, 600 epochs, the learning rate is set to be 0.001, and an Adam optimizer is used for optimizing parameters.
The invention provides a roadside end pedestrian track generation algorithm based on an confrontation generation network, which generates a multi-mode predicted track by utilizing a social attention mechanism and a pedestrian track latent variable; according to the method, the capabilities of the generator and the discriminator are continuously optimized through the confrontation generation training of the trajectory generator and the discriminator, and the accuracy of the trajectory generated by the generator is improved; the invention provides a social attention mechanism based on head orientation, which obtains the head orientation of a pedestrian through the last speed direction of the pedestrian, calculates the cosine value of an included angle between the pedestrians according to the head orientation information, optimizes the output of the social attention mechanism by using the calculated angle information and converges the output through a maximum pooling layer; the invention provides a new latent variable generation method, which is characterized in that two feedforward neural networks are used for learning latent variables from pedestrian historical tracks and observation tracks respectively, the input of a latent variable generator comprises position, speed and acceleration, and the distribution of three types of latent variables is generated from the three types of input respectively; the three types of latent variable distributions are combined with Gaussian random noise to generate multi-modal output and maintain the capability of processing uncertain input in the future.
In light of the foregoing description of the preferred embodiment of the present invention, many modifications and variations will be apparent to those skilled in the art without departing from the spirit and scope of the invention. The technical scope of the present invention is not limited to the content of the specification, and must be determined according to the scope of the claims.

Claims (7)

1.一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：包括如下步骤：1. a roadside pedestrian trajectory prediction algorithm based on confrontation generating network, is characterized in that: comprise the steps: S10：使用编码器编码输入轨迹；S10: use the encoder to encode the input track; S20：利用行人头部朝向计算行人社会注意力；S20: Calculate the pedestrian's social attention by using the pedestrian's head orientation; S30：应用潜在变量预测器来生成可预测的潜在变量分布；S30: Apply a latent variable predictor to generate a predictable latent variable distribution; S40：生成预测的行人未来轨迹；S40: Generate a predicted pedestrian future trajectory; S50：使用判别器来优化生成器生成的行人轨迹；S50: Use the discriminator to optimize the pedestrian trajectory generated by the generator; 所述步骤S30包括如下步骤：The step S30 includes the following steps: S31：设计潜变量预测器；S31: Design a latent variable predictor; S32：使用潜变量预测器预测行人潜在的变量分布。S32: Use a latent variable predictor to predict the latent variable distribution of pedestrians. 2.如权利要求1所述的一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：在步骤S31中：潜变量预测器由两个前馈神经网络组成定义如下：2. a kind of roadside pedestrian trajectory prediction algorithm based on confrontation generating network as claimed in claim 1, is characterized in that: in step S31: latent variable predictor is formed by two feedforward neural networks and is defined as follows:
Figure FDA0002764619680000011
Figure FDA0002764619680000011
Figure FDA0002764619680000012
Figure FDA0002764619680000012
 其中Ψ(·)和
Figure FDA0002764619680000013
是前馈神经网络，
Figure FDA0002764619680000014
和
Figure FDA0002764619680000015
分别是这两个前馈神经网络的参数，
Figure FDA0002764619680000016
和
Figure FDA0002764619680000017
是潜变量预测器第k类输入。where Ψ( ) and
Figure FDA0002764619680000013
is a feedforward neural network,
Figure FDA0002764619680000014
and
Figure FDA0002764619680000015
are the parameters of the two feedforward neural networks, respectively,
Figure FDA0002764619680000016
and
Figure FDA0002764619680000017
is the k-th class input of the latent variable predictor. 3.如权利要求2所述的一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：在步骤S32中：k＝1、2、3、分别表示行人位置、速度、加速度，位置揭示了潜在场景的布局，速度反映了不同行人的运动模式，加速度表明了行人运动激烈程度；潜在变量预测器从这三种输入中估计出三种变量的潜在分布；高斯随机噪声被用于产生多模态输出，最后把这三种潜变量分布和高斯随机噪声融合在一起，最终构成训练阶段的潜变量分布参数；3. a kind of roadside pedestrian trajectory prediction algorithm based on confrontation generation network as claimed in claim 2, it is characterized in that: in step S32: k=1,2,3, respectively represent pedestrian position, speed, acceleration, The location reveals the layout of the latent scene, the velocity reflects the movement patterns of different pedestrians, and the acceleration indicates the intensity of the pedestrian movement; the latent variable predictor estimates the latent distribution of the three variables from these three inputs; Gaussian random noise is used for Generate multi-modal output, and finally fuse the three latent variable distributions with Gaussian random noise, and finally constitute the latent variable distribution parameters in the training phase; 在测试阶段潜在变量预测器从行人观测轨迹中预测潜在变量分布，潜在变量预测器输入的是行人的位置、速度、加速度信息，潜在变量预测器可以从这三类输入中分别预测出行人的位置、速度、加速度的潜在变量分布，把这三种潜在变量和高斯随机噪声结合形成最终的潜在变量输出到轨迹生成器中；In the test phase, the latent variable predictor predicts the distribution of latent variables from the pedestrian observation trajectories. The latent variable predictor inputs the pedestrian's position, speed, and acceleration information. The latent variable predictor can predict the pedestrian's position from these three types of inputs. , the latent variable distribution of velocity and acceleration, and combine these three latent variables with Gaussian random noise to form the final latent variable output to the trajectory generator; 在训练过程中潜在变量损失函数用来衡量观测轨迹的潜在变量分布和真实轨迹潜在变量分布之间的差距，使用KL散度来计算这种误差，公式如下：In the training process, the latent variable loss function is used to measure the gap between the latent variable distribution of the observed trajectory and the true trajectory latent variable distribution. The KL divergence is used to calculate this error, and the formula is as follows:
Figure FDA0002764619680000021
Figure FDA0002764619680000021
 其中
Figure FDA0002764619680000022
和
Figure FDA0002764619680000023
分别表示观测轨迹的潜变量分布和真实轨迹的潜变量分布。in
Figure FDA0002764619680000022
and
Figure FDA0002764619680000023
represent the latent variable distribution of the observed trajectory and the latent variable distribution of the true trajectory, respectively. 4.如权利要求1所述的一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：所述步骤1具体包括如下步骤：4. a kind of roadside pedestrian trajectory prediction algorithm based on confrontation generation network as claimed in claim 1, is characterized in that: described step 1 specifically comprises the following steps: S11：对输入轨迹数据进行处理：输入轨迹是一系列时间序列的轨迹点
Figure FDA0002764619680000024
其中
Figure FDA0002764619680000025
是目标i在t时刻的位置坐标；把每条轨迹在不同时刻的位置坐标送入编码网络中；S11: Process the input trajectory data: the input trajectory is a series of trajectory points in a time series
Figure FDA0002764619680000024
in
Figure FDA0002764619680000025
is the position coordinate of target i at time t; the position coordinates of each trajectory at different times are sent into the encoding network; S12：使用一个单层的多层感知机将两维的位置信息转变成固定长度的多维向量
Figure FDA0002764619680000026
多层感知机的定义如下：S12: Use a single-layer multilayer perceptron to convert two-dimensional position information into fixed-length multi-dimensional vectors
Figure FDA0002764619680000026
The definition of a multilayer perceptron is as follows:
Figure FDA0002764619680000027
Figure FDA0002764619680000027
 其中φ(·)是使用ReLU非线性激活函数的多层感知机，Wee是多层感知机的参数；where φ( ) is the multilayer perceptron using ReLU nonlinear activation function, and Wee is the parameter of the multilayer perceptron; S13：将多维向量送入基于长短期记忆网络的编码器中，生成行人运动的隐状态
Figure FDA0002764619680000028
编码器长短期记忆网络(LSTM)定义如下：S13: Send the multi-dimensional vector into the encoder based on long short-term memory network to generate the hidden state of pedestrian movement
Figure FDA0002764619680000028
The encoder long short-term memory network (LSTM) is defined as follows:
Figure FDA0002764619680000029
Figure FDA0002764619680000029
 其中LSTM(·)是长短期记忆网络，Wencoder是编码器长短期记忆网络的参数，参数可在同一场景下的所有行人之间共享。where LSTM( ) is the long short-term memory network, W encoder is the parameter of the encoder long short-term memory network, and the parameters can be shared among all pedestrians in the same scene. 5.如权利要求1所述的一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：所述步骤2具体包括如下步骤：5. a kind of roadside pedestrian trajectory prediction algorithm based on confrontation generation network as claimed in claim 1, is characterized in that: described step 2 specifically comprises the following steps: S21：计算行人之间的方位夹角：将行人最后位置的速度作为行人未来的速度，把行人最后位置的速度的方向作为头部朝向和轨迹运动方向，利用所有行人头部朝向来计算行人之间方位角的余弦值如下：S21: Calculate the azimuth angle between pedestrians: take the speed of the last position of the pedestrian as the future speed of the pedestrian, take the direction of the speed of the last position of the pedestrian as the head orientation and the trajectory movement direction, and use the head orientation of all pedestrians to calculate the distance between the pedestrians The cosine of the azimuth angle is as follows:
Figure FDA0002764619680000031
Figure FDA0002764619680000031
 其中n是同一场景中所有行人的个数，bij表示行人i和行人j之间的夹角；where n is the number of all pedestrians in the same scene, and b ij represents the angle between pedestrian i and pedestrian j; S22：设计注意力机制：根据行人之间的方位夹角余弦值设计出软注意力机制和硬注意力机制；一个行人对另一个行人的影响随着他们之间方位角余弦值增加而减小；硬注意力机制用一个与cos(β)形状相同的矩阵HA表示，HA中每一个元素hij的值都设置为0或1，当行人之间的方位夹角余弦值大于预设阈值0.2时，对应的注意力权重hij为1，当行人之间的方位夹角余弦值小于预设阈值0.2时，对应的注意力权重hij为0；软注意力机制与硬注意力机制通过阈值来计算注意力权重不同；软注意力机制自适应的计算行人之间的相关性，软注意力机制权重SA的计算公式如下:S22: Design an attention mechanism: Design a soft attention mechanism and a hard attention mechanism according to the cosine value of the azimuth angle between pedestrians; the influence of one pedestrian on another pedestrian decreases as the cosine value of the azimuth angle between them increases ; The hard attention mechanism is represented by a matrix H A with the same shape as cos(β), and the value of each element h ij in H A is set to 0 or 1. When the cosine value of the azimuth angle between pedestrians is greater than the preset When the threshold is 0.2, the corresponding attention weight h ij is 1, and when the cosine value of the azimuth angle between pedestrians is less than the preset threshold 0.2, the corresponding attention weight h ij is 0; soft attention mechanism and hard attention mechanism Different attention weights are calculated through the threshold; the soft attention mechanism adaptively calculates the correlation between pedestrians, and the calculation formula of the soft attention mechanism weight S A is as follows:
Figure FDA0002764619680000032
Figure FDA0002764619680000032
 其中δ(·）表示sigmoid激活函数，
Figure FDA0002764619680000033
表示1x1卷积层；where δ( ) represents the sigmoid activation function,
Figure FDA0002764619680000033
Represents a 1x1 convolutional layer; 软硬注意力机制作用在第二个多层感知机输出上，使用软注意力机制和硬注意力机制来优化第二个多层感知机的输出，通过最大池化层来汇聚注意力机制得到输出
Figure FDA0002764619680000034
The soft and hard attention mechanism acts on the output of the second multi-layer perceptron. The soft-attention mechanism and the hard-attention mechanism are used to optimize the output of the second multi-layer perceptron, and the maximum pooling layer is used to converge the attention mechanism to obtain output
Figure FDA0002764619680000034
 6.如权利要求1所述的一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：所述步骤4具体包括如下步骤：6. a kind of roadside pedestrian trajectory prediction algorithm based on confrontation generation network as claimed in claim 1 is characterized in that: described step 4 specifically comprises the following steps: S41：将社会注意力模块的输出
Figure FDA0002764619680000041
和潜变量预测器的输出
Figure FDA0002764619680000042
与行人运动隐状态
Figure FDA0002764619680000043
进行拼接S41: Combine the output of the social attention module
Figure FDA0002764619680000041
and the output of the latent variable predictor
Figure FDA0002764619680000042
Hidden state with pedestrian motion
Figure FDA0002764619680000043
splicing S42：拼接结果输入到基于长短期记忆网络的解码器中，得到融合了多种信息的新轨迹隐状态
Figure FDA0002764619680000044
解码器的长短期记忆网络定义如下：S42: The splicing result is input into the decoder based on the long short-term memory network, and a new trajectory hidden state that integrates various information is obtained.
Figure FDA0002764619680000044
The long short-term memory network of the decoder is defined as follows:
Figure FDA0002764619680000045
Figure FDA0002764619680000045
 其中LSTM(·)是长短期记忆网络，Wdecoder是解码器长短期记忆网络的参数，参数可在同一场景下所有行人之间共享；Where LSTM( ) is the long short-term memory network, W decoder is the parameters of the decoder long-term and short-term memory network, and the parameters can be shared among all pedestrians in the same scene; S43：用多层感知机对新的隐状态进行解码得到行人未来轨迹坐标：多层感知机定义如下：S43: Decode the new hidden state with the multilayer perceptron to obtain the coordinates of the pedestrian's future trajectory: The multilayer perceptron is defined as follows:
Figure FDA0002764619680000046
Figure FDA0002764619680000046
 其中γ(·)是使用ReLU非线性激活函数的多层感知机，
Figure FDA0002764619680000047
是行人未来的位置坐标，输出的预测结果为一系列的位置坐标
Figure FDA0002764619680000048
其中Tobs是预测轨迹的长度；本发明采用多模式输出，轨迹生成器一次输出m条轨迹，使用2范数损失函数计算这m条轨迹和真实值之间的偏差，表达式如下：where γ( ) is a multilayer perceptron using ReLU nonlinear activation function,
Figure FDA0002764619680000047
is the future position coordinates of the pedestrian, and the output prediction result is a series of position coordinates
Figure FDA0002764619680000048
Where T obs is the length of the predicted trajectory; the present invention adopts multi-mode output, the trajectory generator outputs m trajectories at a time, and uses the 2-norm loss function to calculate the deviation between the m trajectories and the true value, and the expression is as follows:
Figure FDA0002764619680000049
Figure FDA0002764619680000049
 其中
Figure FDA00027646196800000410
是行人真实轨迹，
Figure FDA00027646196800000411
是第m条生成器预测的行人未来轨迹，在本发明中设置m＝20。in
Figure FDA00027646196800000410
is the true trajectory of pedestrians,
Figure FDA00027646196800000411
is the pedestrian future trajectory predicted by the mth generator, and m=20 is set in the present invention. 7.如权利要求1所述的一种基于对抗生成网络的路侧端行人轨迹预测算法，其特征在于：所述步骤5具体包括如下步骤：7. a kind of roadside pedestrian trajectory prediction algorithm based on confrontation generation network as claimed in claim 1 is characterized in that: described step 5 specifically comprises the following steps: S51：将生成器生成的轨迹和行人真实的轨迹输入到判别器中S51: Input the trajectory generated by the generator and the real trajectory of the pedestrian into the discriminator S52：判别器判别输入轨迹是由生成器生成的还是真实的轨迹：判别器使用一个基于长短期记忆网络的编码器会对真实的轨迹和生成的轨迹进行编码，在编码器输出的隐状态上应用多层感知机来获得分类得分，理想情况下判别器会学习到行人轨迹的社会规则，不符合这个规则的轨迹将被判别器判定为假；S52: The discriminator discriminates whether the input trajectory is generated by the generator or the real trajectory: the discriminator uses an encoder based on a long short-term memory network to encode the real trajectory and the generated trajectory, on the hidden state output by the encoder Apply a multi-layer perceptron to obtain classification scores. Ideally, the discriminator will learn the social rules of pedestrian trajectories, and trajectories that do not meet this rule will be judged as false by the discriminator; 对抗生成训练的损失函数表述如下：The loss function for adversarial generative training is expressed as follows:
Figure FDA0002764619680000051
Figure FDA0002764619680000051
 其中D是判别器，G是生成器，z是潜变量分布参数，x是数据的轨迹数据，
Figure FDA0002764619680000052
是观测轨迹潜变量预测器的第k个输入(位置，速度，加速度)；通过生成器和判别器的博弈训练，生成器最终可以生成类似训练集的，符合社会规则的样本；由于生成器学到是一个和训练集类似的概率分布，每次采样都可以给出不同的合理样本，故可被用来对多种可能性进行预测；where D is the discriminator, G is the generator, z is the latent variable distribution parameter, x is the trajectory data of the data,
Figure FDA0002764619680000052
is the kth input (position, velocity, acceleration) of the observed trajectory latent variable predictor; through the game training of the generator and the discriminator, the generator can finally generate samples similar to the training set and conform to social rules; It is a probability distribution similar to the training set, and each sampling can give different reasonable samples, so it can be used to predict multiple possibilities; 总的损失函数由三部分构成，一部分是对抗生成训练损失函数，一部分是潜变量分布KL散度，一部分是预测值与真实值之间的偏差；总的损失函数加权定义如下：The total loss function consists of three parts, one part is the adversarial generation training loss function, one part is the latent variable distribution KL divergence, and the other part is the deviation between the predicted value and the true value; the total loss function weighting is defined as follows:
Figure FDA0002764619680000053
Figure FDA0002764619680000053
 其中α和β分别设置为1到10之间的数字，具体的数值通过在基准数据集上交叉验证得到；在训练时，迭代训练生成器和判别器，设置批处理大小为64，600个epoch，学习率设置为0.001，使用Adam优化器对参数进行优化。where α and β are respectively set to numbers between 1 and 10, and the specific values are obtained by cross-validation on the benchmark dataset; during training, the generator and discriminator are iteratively trained, and the batch size is set to 64,600 epochs , the learning rate is set to 0.001, and the parameters are optimized using the Adam optimizer.
CN202011229272.0A 2020-11-06 2020-11-06 A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network Withdrawn CN112347923A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011229272.0A CN112347923A (en) 2020-11-06 2020-11-06 A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011229272.0A CN112347923A (en) 2020-11-06 2020-11-06 A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network

Publications (1)

Publication Number Publication Date
CN112347923A true CN112347923A (en) 2021-02-09

Family

ID=74428364

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011229272.0A Withdrawn CN112347923A (en) 2020-11-06 2020-11-06 A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network

Country Status (1)

Country Link
CN (1) CN112347923A (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113156957A (en) * 2021-04-27 2021-07-23 东莞理工学院 Autonomous mobile robot self-supervision learning and navigation method based on confrontation generation network
CN113473682A (en) * 2021-09-01 2021-10-01 启东晶尧光电科技有限公司 Artificial intelligence-based smart city lighting street lamp adjusting method and system
CN113538506A (en) * 2021-07-23 2021-10-22 陕西师范大学 Pedestrian trajectory prediction method based on deep modeling of global dynamic scene information
CN113689470A (en) * 2021-09-02 2021-11-23 重庆大学 Pedestrian motion trajectory prediction method under multi-scene fusion
CN113761395A (en) * 2021-08-12 2021-12-07 清华大学 Trajectory generation model training method, trajectory generation method and device
CN114155270A (en) * 2021-11-10 2022-03-08 南方科技大学 Pedestrian trajectory prediction method, device, device and storage medium
CN114170680A (en) * 2021-11-25 2022-03-11 北京易航远智科技有限公司 Pedestrian trajectory prediction method based on multi-subdivision feature control
CN114511594A (en) * 2022-01-27 2022-05-17 西安交通大学 Multi-mode pedestrian trajectory prediction method and system based on predefined tree
CN114757975A (en) * 2022-04-29 2022-07-15 华南理工大学 Pedestrian Trajectory Prediction Method Based on Transformer and Graph Convolutional Network
CN114898550A (en) * 2022-03-16 2022-08-12 清华大学 Pedestrian trajectory prediction method and system
CN115272712A (en) * 2022-07-26 2022-11-01 北京邮电大学 Pedestrian trajectory prediction method fusing moving target analysis
CN115719479A (en) * 2022-11-30 2023-02-28 电子科技大学长三角研究院(衢州) Track prediction method based on encoder-decoder architecture
CN118710786A (en) * 2024-08-23 2024-09-27 山东云小华数字科技有限公司 A digital human motion generation method and generation system

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113156957A (en) * 2021-04-27 2021-07-23 东莞理工学院 Autonomous mobile robot self-supervision learning and navigation method based on confrontation generation network
CN113156957B (en) * 2021-04-27 2024-06-04 东莞理工学院 Self-supervision learning and navigation method for autonomous mobile robot
CN113538506A (en) * 2021-07-23 2021-10-22 陕西师范大学 Pedestrian trajectory prediction method based on deep modeling of global dynamic scene information
CN113761395A (en) * 2021-08-12 2021-12-07 清华大学 Trajectory generation model training method, trajectory generation method and device
CN113761395B (en) * 2021-08-12 2024-07-26 清华大学 Track generation model training method, track generation method and track generation device
CN113473682A (en) * 2021-09-01 2021-10-01 启东晶尧光电科技有限公司 Artificial intelligence-based smart city lighting street lamp adjusting method and system
CN113473682B (en) * 2021-09-01 2021-11-26 启东晶尧光电科技有限公司 Artificial intelligence-based smart city lighting street lamp adjusting method and system
CN113689470B (en) * 2021-09-02 2023-08-11 重庆大学 A Pedestrian Trajectory Prediction Method Based on Multi-Scene Fusion
CN113689470A (en) * 2021-09-02 2021-11-23 重庆大学 Pedestrian motion trajectory prediction method under multi-scene fusion
CN114155270A (en) * 2021-11-10 2022-03-08 南方科技大学 Pedestrian trajectory prediction method, device, device and storage medium
CN114170680A (en) * 2021-11-25 2022-03-11 北京易航远智科技有限公司 Pedestrian trajectory prediction method based on multi-subdivision feature control
CN114511594A (en) * 2022-01-27 2022-05-17 西安交通大学 Multi-mode pedestrian trajectory prediction method and system based on predefined tree
CN114898550A (en) * 2022-03-16 2022-08-12 清华大学 Pedestrian trajectory prediction method and system
CN114898550B (en) * 2022-03-16 2024-03-19 清华大学 A pedestrian trajectory prediction method and system
CN114757975B (en) * 2022-04-29 2024-04-16 华南理工大学 Pedestrian track prediction method based on transformer and graph convolution network
CN114757975A (en) * 2022-04-29 2022-07-15 华南理工大学 Pedestrian Trajectory Prediction Method Based on Transformer and Graph Convolutional Network
CN115272712A (en) * 2022-07-26 2022-11-01 北京邮电大学 Pedestrian trajectory prediction method fusing moving target analysis
CN115719479A (en) * 2022-11-30 2023-02-28 电子科技大学长三角研究院(衢州) Track prediction method based on encoder-decoder architecture
CN115719479B (en) * 2022-11-30 2024-09-17 电子科技大学长三角研究院(衢州) Track prediction method based on encoder-decoder architecture
CN118710786A (en) * 2024-08-23 2024-09-27 山东云小华数字科技有限公司 A digital human motion generation method and generation system

Similar Documents

Publication Publication Date Title
CN112347923A (en) A Roadside Pedestrian Trajectory Prediction Algorithm Based on Adversarial Generative Network
CN111432989B (en) Artificial enhancement cloud-based robot intelligent framework and related methods
Yang et al. A novel graph-based trajectory predictor with pseudo-oracle
Wang et al. Robust imitation of diverse behaviors
Zhao et al. A spatial-temporal attention model for human trajectory prediction.
CN113795851B (en) Large-Scale Generative Neural Network Models with Inference for Representation Learning Using Adversarial Training
Schaal et al. Real-time robot learning with locally weighted statistical learning
US8990133B1 (en) Apparatus and methods for state-dependent learning in spiking neuron networks
CN112714896A (en) Self-aware visual-text common ground navigation agent
CN112418421B (en) Road side end pedestrian track prediction algorithm based on graph attention self-coding model
CN112119409A (en) Neural network with relational memory
KR20220045215A (en) Gated Attention Neural Network
Sang et al. Human motion prediction based on attention mechanism
Wei et al. Learning motion rules from real data: Neural network for crowd simulation
CN114155270A (en) Pedestrian trajectory prediction method, device, device and storage medium
KR20230119023A (en) Attention neural networks with short-term memory
KR20240057422A (en) Control interactive agents using multi-mode input
Lin et al. Research on UUV obstacle avoiding method based on recurrent neural networks
Tanwani Generative models for learning robot manipulation skills from humans
Dan Spatial-temporal block and LSTM network for pedestrian trajectories prediction
CN119028130A (en) Explainable vehicle trajectory prediction method and system based on weighted graph attention network
Pak et al. Carnet: A dynamic autoencoder for learning latent dynamics in autonomous driving tasks
Wu et al. A novel trajectory generator based on a constrained GAN and a latent variables predictor
Pertsch et al. Keyin: Discovering subgoal structure with keyframe-based video prediction
He et al. An adversarial learned trajectory predictor with knowledge-rich latent variables

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
WW01 Invention patent application withdrawn after publication

Application publication date: 20210209

WW01 Invention patent application withdrawn after publication