자기 주의 증류를 이용한 심층 신경망 기반의 그림자 제거

다국어 초록 (Multilingual Abstract)

Shadow removal plays a key role for the pre-processing of image processing techniques such as object tracking and detection. With the advances of image recognition based on deep convolution neural networks, researches for shadow removal have been actively conducted. In this paper, we propose a novel method for shadow removal, which utilizes self attention distillation to extract semantic features. The proposed method gradually refines results of shadow detection, which are extracted from each layer of the proposed network, via top-down distillation. Specifically, the training procedure can be efficiently performed by learning the contextual information for shadow removal without shadow masks. Experimental results on various datasets show the effectiveness of the proposed method for shadow removal under real world environments.

참고문헌 (Reference)

1 K. Simonyan, "Very deep convolutional networks for large-scale image recognition" 1-14, 2015

2 J. Wang, "Stacked conditional generative adversarial networks for jointly learning shadow detection and shadow removal" 1788-1797, 2018

3 Y. Zhang, "Residual dense network for image super-resolution" 2472-2481, 2018

4 F. Wang, "Residual attention network for image classification" 3156-3164, 2017

5 S. Zagoruyko, "Paying more attention to attention:improving the performance of convolutional neural networks via attention transfer" 1-13, 2017

6 R. Guo, "Paired regions for shadow detection and removal" 35 (35): 2956-2967, 2013

7 G. Finlayson, "On the removal of shadows from images" 28 (28): 59-68, 2006

8 H. Gong, "Interactive shadow removal and ground truth for variable scene categories" 1-11, 2014

9 A. Krizhevsky, "ImageNet classification with deep convolutional neural networks" 1097-1105, 2017

10 G. Hinton, "Distilling the knowledge in a neural network" 1-9, 2015

11 X. Hu, "Direction-aware spatial context features for shadow detection and removal" 42 (42): 2795-2808, 2019

12 L. Qu, "DeshadowNet: A multi-context embedding deep network for shadow removal" 4067-4075, 2017

13 A. Paszke, "Automatic differentiation in pytorch" 1-4, 2017

14 S. Xie, "Aggregated residual transformations for deep neural networks" 1493-1500, 2017