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

Towards Effective and Efficient 3D V...

Hong, Ziyang,

Linked to FindBook

Google Book

Amazon

博客來

Towards Effective and Efficient 3D Visual Perception /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Towards Effective and Efficient 3D Visual Perception // Ziyang Hong.
Author:	Hong, Ziyang,
Description:	1 electronic resource (102 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-05, Section: B.
Contained By:	Dissertations Abstracts International86-05B.
Subject:	Computer engineering. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31718291
ISBN:	9798346544227

Towards Effective and Efficient 3D Visual Perception /
Hong, Ziyang,

Towards Effective and Efficient 3D Visual Perception /Ziyang Hong. - 1 electronic resource (102 pages)

Source: Dissertations Abstracts International, Volume: 86-05, Section: B.

We are witnessing developments in artificial intelligence (AI), where the most relevant parts to us are embodied AI systems with robotic agents interacting with human users in real-world environments. To serve in the embodied AI system, the agent robots need visual intelligence, namely the three-dimensional (3D) perception of the surrounding envi ronment. This thesis is dedicated to enabling real-time low-cost edge deployment of visual perception which is advocated by us as the future solution for embodied AI applications.First, an end-to-end deep neural network pipeline for machinery visual perception, CDRNet, is proposed. It jointly perceives a 3D scene's geometry structure and semantic labels. While conventional volumetric approaches for 3D perception tend to focus on the global coherence of their reconstructions, which leads to a lack of local geometric detail, CDRNet leverages the latent geometric prior knowledge in 2D image features by explicit depth prediction and anchored feature generation, to refine the occupancy learning in TSDF volume.Besides, we find that this cross-dimensional feature refinement methodology can also be adopted for the semantic segmentation task by utilizing semantic priors, to extract both 3D mesh and 3D semantic labeling in real-time. Beyond public datasets for test ing, we further implement a real-time messaging system to support these aforementioned perception tasks in real-life scenarios.Finally, a software-hardware co-optimization system, Efficient-Grad is proposed to enable the online AI model fine-tuning. It improves both throughput and energy saving with negligible accuracy degradation during model training for deep convolutional neural networks, by utilizing sparsity and asymmetry residing in the gradients for conventional back propagation. Furthermore, the dedicated hardware architecture for sparsity utiliza tion and efficient data movement is optimized to support the Efficient-Grad algorithm in a scalable manner, which leads to its superiority in terms of energy efficiency.

English

ISBN: 9798346544227Subjects--Topical Terms:

621879
Computer engineering.

Towards Effective and Efficient 3D Visual Perception /
LDR:03292nmm a22003733i 4500 001 2435940
005 20251014050545.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798346544227
035 $a (MiAaPQD)AAI31718291
035 $a AAI31718291
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Hong, Ziyang, $e author. $3 3826429
245 1 0 $a Towards Effective and Efficient 3D Visual Perception / $c Ziyang Hong.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (102 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-05, Section: B.
500 $a Advisors: Yue, Chik Patrick.
502 $b Ph.D. $c Hong Kong University of Science and Technology (Hong Kong) $d 2024.
520 $a We are witnessing developments in artificial intelligence (AI), where the most relevant parts to us are embodied AI systems with robotic agents interacting with human users in real-world environments. To serve in the embodied AI system, the agent robots need visual intelligence, namely the three-dimensional (3D) perception of the surrounding envi ronment. This thesis is dedicated to enabling real-time low-cost edge deployment of visual perception which is advocated by us as the future solution for embodied AI applications.First, an end-to-end deep neural network pipeline for machinery visual perception, CDRNet, is proposed. It jointly perceives a 3D scene's geometry structure and semantic labels. While conventional volumetric approaches for 3D perception tend to focus on the global coherence of their reconstructions, which leads to a lack of local geometric detail, CDRNet leverages the latent geometric prior knowledge in 2D image features by explicit depth prediction and anchored feature generation, to refine the occupancy learning in TSDF volume.Besides, we find that this cross-dimensional feature refinement methodology can also be adopted for the semantic segmentation task by utilizing semantic priors, to extract both 3D mesh and 3D semantic labeling in real-time. Beyond public datasets for test ing, we further implement a real-time messaging system to support these aforementioned perception tasks in real-life scenarios.Finally, a software-hardware co-optimization system, Efficient-Grad is proposed to enable the online AI model fine-tuning. It improves both throughput and energy saving with negligible accuracy degradation during model training for deep convolutional neural networks, by utilizing sparsity and asymmetry residing in the gradients for conventional back propagation. Furthermore, the dedicated hardware architecture for sparsity utiliza tion and efficient data movement is optimized to support the Efficient-Grad algorithm in a scalable manner, which leads to its superiority in terms of energy efficiency.
546 $a English
590 $a School code: 1223
650 4 $a Computer engineering. $3 621879
650 4 $a Energy consumption. $3 631630
650 4 $a Neural networks. $3 677449
650 4 $a Energy. $3 876794
690 $a 0464
690 $a 0800
690 $a 0791
710 2 $a Hong Kong University of Science and Technology (Hong Kong). $e degree granting institution. $3 3814653
720 1 $a Yue, Chik Patrick $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-05B.
790 $a 1223
791 $a Ph.D.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31718291