東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Low-Shot Learning for the Semantic S...

Kemker, Ronald.

FindBook

Google Book

Amazon

博客來

Low-Shot Learning for the Semantic Segmentation of Remote Sensing Imagery.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Low-Shot Learning for the Semantic Segmentation of Remote Sensing Imagery./
作者:	Kemker, Ronald.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2018,
面頁冊數:	137 p.
附註:	Source: Dissertation Abstracts International, Volume: 80-01(E), Section: B.
Contained By:	Dissertation Abstracts International80-01B(E).
標題:	Remote sensing. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10931043
ISBN:	9780438303188

Low-Shot Learning for the Semantic Segmentation of Remote Sensing Imagery.
Kemker, Ronald.

Low-Shot Learning for the Semantic Segmentation of Remote Sensing Imagery. - Ann Arbor : ProQuest Dissertations & Theses, 2018 - 137 p.

Source: Dissertation Abstracts International, Volume: 80-01(E), Section: B.

Thesis (Ph.D.)--Rochester Institute of Technology, 2018.

Deep-learning frameworks have made remarkable progress thanks to the creation of large annotated datasets such as ImageNet, which has over one million training images. Although this works well for color (RGB) imagery, labeled datasets for other sensor modalities (e.g., multispectral and hyperspectral) are minuscule in comparison. This is because annotated datasets are expensive and man-power intensive to complete; and since this would be impractical to accomplish for each type of sensor, current state-of-the-art approaches in computer vision are not ideal for remote sensing problems. The shortage of annotated remote sensing imagery beyond the visual spectrum has forced researchers to embrace unsupervised feature extracting frameworks. These features are learned on a per-image basis, so they tend to not generalize well across other datasets. In this dissertation, we propose three new strategies for learning feature extracting frameworks with only a small quantity of annotated image data; including 1) self-taught feature learning, 2) domain adaptation with synthetic imagery, and 3) semi-supervised classification. "Self-taught" feature learning frameworks are trained with large quantities of unlabeled imagery, and then these networks extract spatial-spectral features from annotated data for supervised classification. Synthetic remote sensing imagery can be used to boot-strap a deep convolutional neural network, and then we can fine-tune the network with real imagery. Semi-supervised classifiers prevent overfitting by jointly optimizing the supervised classification task along side one or more unsupervised learning tasks (i.e., reconstruction). Although obtaining large quantities of annotated image data would be ideal, our work shows that we can make due with less cost-prohibitive methods which are more practical to the end-user.

ISBN: 9780438303188Subjects--Topical Terms:

535394
Remote sensing.

Low-Shot Learning for the Semantic Segmentation of Remote Sensing Imagery.
LDR:02813nmm a2200301 4500 001 2201346
005 20190429062347.5
008 201008s2018 ||||||||||||||||| ||eng d
020 $a 9780438303188
035 $a (MiAaPQ)AAI10931043
035 $a (MiAaPQ)rit:13055
035 $a AAI10931043
040 $a MiAaPQ $c MiAaPQ
100 1 $a Kemker, Ronald. $3 3428070
245 1 0 $a Low-Shot Learning for the Semantic Segmentation of Remote Sensing Imagery.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2018
300 $a 137 p.
500 $a Source: Dissertation Abstracts International, Volume: 80-01(E), Section: B.
500 $a Adviser: Christopher Kanan.
502 $a Thesis (Ph.D.)--Rochester Institute of Technology, 2018.
520 $a Deep-learning frameworks have made remarkable progress thanks to the creation of large annotated datasets such as ImageNet, which has over one million training images. Although this works well for color (RGB) imagery, labeled datasets for other sensor modalities (e.g., multispectral and hyperspectral) are minuscule in comparison. This is because annotated datasets are expensive and man-power intensive to complete; and since this would be impractical to accomplish for each type of sensor, current state-of-the-art approaches in computer vision are not ideal for remote sensing problems. The shortage of annotated remote sensing imagery beyond the visual spectrum has forced researchers to embrace unsupervised feature extracting frameworks. These features are learned on a per-image basis, so they tend to not generalize well across other datasets. In this dissertation, we propose three new strategies for learning feature extracting frameworks with only a small quantity of annotated image data; including 1) self-taught feature learning, 2) domain adaptation with synthetic imagery, and 3) semi-supervised classification. "Self-taught" feature learning frameworks are trained with large quantities of unlabeled imagery, and then these networks extract spatial-spectral features from annotated data for supervised classification. Synthetic remote sensing imagery can be used to boot-strap a deep convolutional neural network, and then we can fine-tune the network with real imagery. Semi-supervised classifiers prevent overfitting by jointly optimizing the supervised classification task along side one or more unsupervised learning tasks (i.e., reconstruction). Although obtaining large quantities of annotated image data would be ideal, our work shows that we can make due with less cost-prohibitive methods which are more practical to the end-user.
590 $a School code: 0465.
650 4 $a Remote sensing. $3 535394
650 4 $a Artificial intelligence. $3 516317
690 $a 0799
690 $a 0800
710 2 $a Rochester Institute of Technology. $b Imaging Science. $3 1019498
773 0 $t Dissertation Abstracts International $g 80-01B(E).
790 $a 0465
791 $a Ph.D.
792 $a 2018
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10931043