"Dilated cnn"의 두 판 사이의 차이

2017년 3월 24일 (금) 00:37 판

dense prediction : "The goal is to compute a discrete or continuous label for each pixel in the image.”
- good example is semantic segmentation
  - multi-scale contextual reasoning? (He et al., 2004; Galleguillos & Belongie, 2010).
- ref. Sercu, Tom, and Vaibhava Goel. "Dense Prediction on Sequences with Time-Dilated Convolutions for Speech Recognition." arXiv preprint arXiv:1611.09288 (2016).
'The familiar discrete convolution is simply the 1-dilated convolution. ’
ref. Dilated Convolutions and Kronecker Factored Convolutions ★

@@ 1번째 줄: / 1번째 줄: @@
 ==[https://arxiv.org/pdf/1511.07122.pdf Yu, Fisher, and Vladlen Koltun. "Multi-scale context aggregation by dilated convolutions." arXiv preprint arXiv:1511.07122 (2015).]==
-* dense prediction?
+* dense prediction : "The goal is to compute a discrete or continuous label for each pixel in the image.”
-** [https://arxiv.org/abs/1611.09288v2 Sercu, Tom, and Vaibhava Goel. "Dense Prediction on Sequences with Time-Dilated Convolutions for Speech Recognition." arXiv preprint arXiv:1611.09288 (2016).]
+** good example is semantic segmentation
+*** multi-scale contextual reasoning? (He et al., 2004; Galleguillos & Belongie, 2010).
+** ref. [https://arxiv.org/abs/1611.09288v2 Sercu, Tom, and Vaibhava Goel. "Dense Prediction on Sequences with Time-Dilated Convolutions for Speech Recognition." arXiv preprint arXiv:1611.09288 (2016).]
+* '''The familiar discrete convolution is simply the 1-dilated convolution. ''’
+* ref. [http://www.inference.vc/dilated-convolutions-and-kronecker-factorisation/ Dilated Convolutions and Kronecker Factored Convolutions] ★
+==etc==
+* CRF : [https://arxiv.org/abs/1412.7062 Chen, Liang-Chieh, et al. "Semantic image segmentation with deep convolutional nets and fully connected crfs." arXiv preprint arXiv:1412.7062 (2014).]