東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Analysis of Continuous Learning Mode...

Norman, Kendal Graham.

Linked to FindBook

Google Book

Amazon

博客來

Analysis of Continuous Learning Models For Trajectory Representation.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Analysis of Continuous Learning Models For Trajectory Representation./
Author:	Norman, Kendal Graham.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2023,
Description:	52 p.
Notes:	Source: Masters Abstracts International, Volume: 84-12.
Contained By:	Masters Abstracts International84-12.
Subject:	Success. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30501577
ISBN:	9798379667160

Analysis of Continuous Learning Models For Trajectory Representation.
Norman, Kendal Graham.

Analysis of Continuous Learning Models For Trajectory Representation. - Ann Arbor : ProQuest Dissertations & Theses, 2023 - 52 p.

Source: Masters Abstracts International, Volume: 84-12.

Thesis (M.Sc.)--Purdue University, 2023.

Trajectory planning is a field with widespread utility, and imitation learning pipelines show promise as an accessible training method for trajectory planning. MPNet is the state of the art for imitation learning with respect to success rates. MPNet has two general components to its runtime: a neural network predicts the location of the next anchor point in a trajectory, and then planning infrastructure applies sampling-based techniques to produce near-optimal, collision-less paths. This distinction between the two parts of MPNet prompts investigation into the role of the neural architectures in the Neural Motion Planning pipeline, to discover where improvements can be made. This thesis seeks to explore the importance of neural architecture choice by removing the planning structures, and comparing MPNet's feedforward anchor point predictor with that of a continuous model trained to output a continuous trajectory from start to goal. A new state of the art model in continuous learning is the Neural Flow model. As a continuous model, it possess a low standard deviation runtime which can be properly leveraged in the absence of planning infrastructure. Neural Flows also output smooth, continuous trajectory curves that serve to reduce noisy path outputs in the absence of lazy vertex contraction. This project analyzes the performance of MPNet, Resnet Flow, and Coupling Flow models when sampling-based planning tools such as dropout, lazy vertex contraction, and replanning are removed. Each neural planner is trained end-to-end in an imitation learning pipeline utilizing a simple feedforward encoder, a CNN-based encoder, and a Pointnet encoder to encode the environment, for purposes of comparison. Results indicate that performance is competitive, with Neural Flows slightly outperforming MPNet's success rates on our reduced dataset in Simple2D, and being slighty outperformed by MPNet with respect to collision penetration distance in our UR5 Cubby test suite. These results indicate that continuous models can compete with the performance of anchor point predictor models when sampling-based planning techniques are not applied. Neural Flow models also have other benefits that anchor point predictors do not, like continuity guarantees, the ability to select a proportional location in a trajectory to output, and smoothness.

ISBN: 9798379667160Subjects--Topical Terms:

518195
Success.

Analysis of Continuous Learning Models For Trajectory Representation.
LDR:03357nmm a2200349 4500 001 2403739
005 20241125080204.5
006 m o d
007 cr#unu||||||||
008 251215s2023 ||||||||||||||||| ||eng d
020 $a 9798379667160
035 $a (MiAaPQ)AAI30501577
035 $a (MiAaPQ)Purdue22681492
035 $a AAI30501577
040 $a MiAaPQ $c MiAaPQ
100 1 $a Norman, Kendal Graham. $3 3774019
245 1 0 $a Analysis of Continuous Learning Models For Trajectory Representation.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2023
300 $a 52 p.
500 $a Source: Masters Abstracts International, Volume: 84-12.
500 $a Advisor: Qureshi, Ahmed H.
502 $a Thesis (M.Sc.)--Purdue University, 2023.
520 $a Trajectory planning is a field with widespread utility, and imitation learning pipelines show promise as an accessible training method for trajectory planning. MPNet is the state of the art for imitation learning with respect to success rates. MPNet has two general components to its runtime: a neural network predicts the location of the next anchor point in a trajectory, and then planning infrastructure applies sampling-based techniques to produce near-optimal, collision-less paths. This distinction between the two parts of MPNet prompts investigation into the role of the neural architectures in the Neural Motion Planning pipeline, to discover where improvements can be made. This thesis seeks to explore the importance of neural architecture choice by removing the planning structures, and comparing MPNet's feedforward anchor point predictor with that of a continuous model trained to output a continuous trajectory from start to goal. A new state of the art model in continuous learning is the Neural Flow model. As a continuous model, it possess a low standard deviation runtime which can be properly leveraged in the absence of planning infrastructure. Neural Flows also output smooth, continuous trajectory curves that serve to reduce noisy path outputs in the absence of lazy vertex contraction. This project analyzes the performance of MPNet, Resnet Flow, and Coupling Flow models when sampling-based planning tools such as dropout, lazy vertex contraction, and replanning are removed. Each neural planner is trained end-to-end in an imitation learning pipeline utilizing a simple feedforward encoder, a CNN-based encoder, and a Pointnet encoder to encode the environment, for purposes of comparison. Results indicate that performance is competitive, with Neural Flows slightly outperforming MPNet's success rates on our reduced dataset in Simple2D, and being slighty outperformed by MPNet with respect to collision penetration distance in our UR5 Cubby test suite. These results indicate that continuous models can compete with the performance of anchor point predictor models when sampling-based planning techniques are not applied. Neural Flow models also have other benefits that anchor point predictors do not, like continuity guarantees, the ability to select a proportional location in a trajectory to output, and smoothness.
590 $a School code: 0183.
650 4 $a Success. $3 518195
650 4 $a Planning. $3 552734
650 4 $a Ordinary differential equations. $3 3563004
650 4 $a Neural networks. $3 677449
650 4 $a Civil engineering. $3 860360
650 4 $a Mathematics. $3 515831
690 $a 0800
690 $a 0543
690 $a 0501
690 $a 0405
710 2 $a Purdue University. $3 1017663
773 0 $t Masters Abstracts International $g 84-12.
790 $a 0183
791 $a M.Sc.
792 $a 2023
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30501577