國立虎尾科技大學 |

Learning, Moving, and Predicting with Global Motion Representations.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Learning, Moving, and Predicting with Global Motion Representations./
作者:	Jaegle, Andrew.
面頁冊數:	1 online resource (135 pages)
附註:	Source: Dissertation Abstracts International, Volume: 79-10(E), Section: B.
Contained By:	Dissertation Abstracts International79-10B(E).
標題:	Neurosciences. -
電子資源:	click for full text (PQDT)
ISBN:	9780438037113

Learning, Moving, and Predicting with Global Motion Representations.
Jaegle, Andrew.

Learning, Moving, and Predicting with Global Motion Representations. - 1 online resource (135 pages)

Source: Dissertation Abstracts International, Volume: 79-10(E), Section: B.

Thesis (Ph.D.)--University of Pennsylvania, 2018.

Includes bibliographical references

In order to effectively respond to and influence the world they inhabit, animals and other intelligent agents must understand and predict the state of the world and its dynamics. An agent that can characterize how the world moves is better equipped to engage it. Current methods of motion computation rely on local representations of motion (such as optical flow) or simple, rigid global representations (such as camera motion). These methods are useful, but they are difficult to estimate reliably and limited in their applicability to real-world settings, where agents frequently must reason about complex, highly nonrigid motion over long time horizons. In this dissertation, I present methods developed with the goal of building more flexible and powerful notions of motion needed by agents facing the challenges of a dynamic, nonrigid world. This work is organized around a view of motion as a global phenomenon that is not adequately addressed by local or low-level descriptions, but that is best understood when analyzed at the level of whole images and scenes. I develop methods to: (i) robustly estimate camera motion from noisy optical flow estimates by exploiting the global, statistical relationship between the optical flow field and camera motion under projective geometry; (ii) learn representations of visual motion directly from unlabeled image sequences using learning rules derived from a formulation of image transformation in terms of its group properties; (iii) predict future frames of a video by learning a joint representation of the instantaneous state of the visual world and its motion, using a view of motion as transformations of world state. I situate this work in the broader context of ongoing computational and biological investigations into the problem of estimating motion for intelligent perception and action.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780438037113Subjects--Topical Terms:

593561
Neurosciences.
Index Terms--Genre/Form:

554714
Electronic books.

Learning, Moving, and Predicting with Global Motion Representations.
LDR:03099ntm a2200349Ki 4500 001 916902
005 20180928111503.5
006 m o u
007 cr mn||||a|a||
008 190606s2018 xx obm 000 0 eng d
020 $a 9780438037113
035 $a (MiAaPQ)AAI10808343
035 $a (MiAaPQ)upenngdas:13234
035 $a AAI10808343
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Jaegle, Andrew. $3 1190768
245 1 0 $a Learning, Moving, and Predicting with Global Motion Representations.
264 0 $c 2018
300 $a 1 online resource (135 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: Dissertation Abstracts International, Volume: 79-10(E), Section: B.
500 $a Adviser: Kostas Daniilidis.
502 $a Thesis (Ph.D.)--University of Pennsylvania, 2018.
504 $a Includes bibliographical references
520 $a In order to effectively respond to and influence the world they inhabit, animals and other intelligent agents must understand and predict the state of the world and its dynamics. An agent that can characterize how the world moves is better equipped to engage it. Current methods of motion computation rely on local representations of motion (such as optical flow) or simple, rigid global representations (such as camera motion). These methods are useful, but they are difficult to estimate reliably and limited in their applicability to real-world settings, where agents frequently must reason about complex, highly nonrigid motion over long time horizons. In this dissertation, I present methods developed with the goal of building more flexible and powerful notions of motion needed by agents facing the challenges of a dynamic, nonrigid world. This work is organized around a view of motion as a global phenomenon that is not adequately addressed by local or low-level descriptions, but that is best understood when analyzed at the level of whole images and scenes. I develop methods to: (i) robustly estimate camera motion from noisy optical flow estimates by exploiting the global, statistical relationship between the optical flow field and camera motion under projective geometry; (ii) learn representations of visual motion directly from unlabeled image sequences using learning rules derived from a formulation of image transformation in terms of its group properties; (iii) predict future frames of a video by learning a joint representation of the instantaneous state of the visual world and its motion, using a view of motion as transformations of world state. I situate this work in the broader context of ongoing computational and biological investigations into the problem of estimating motion for intelligent perception and action.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Neurosciences. $3 593561
650 4 $a Computer science. $3 573171
650 4 $a Artificial intelligence. $3 559380
655 7 $a Electronic books. $2 local $3 554714
690 $a 0317
690 $a 0984
690 $a 0800
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a University of Pennsylvania. $b Neuroscience. $3 1190675
773 0 $t Dissertation Abstracts International $g 79-10B(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10808343 $z click for full text (PQDT)