iMTE

논문 제목 : Smooth Grad-CAM++: An Enhanced Inference Level Visualization Technique for Deep Convolutional Neural Network Models 논문 주소 : arxiv.org/abs/1908.01224 Smooth Grad-CAM++: An Enhanced Inference Level Visualization Technique for Deep Convolutional Neural Network Models Gaining insight into how deep convolutional neural network models perform image classification and how to explain their outpu..

논문 제목 : Grad-CAM++: Generalized Gradient-based Visual Explanations for Deep Convolutional Networks 논문 주소 : arxiv.org/pdf/1710.11063.pdf IEEE WACV (2018, ieeexplore.ieee.org/document/8354201)에 나온 논문을 바탕으로 이해하고 내용을 작성한다. arixv에서 나온 버전이 좀 더 extended version임으로 Grad-CAM++에 더 깊은 이해를 위해서는 extended version을 읽는 것을 추천한다. 주요 내용 : 1) Deep models은 "black box"로서 internal function을 이해하는데에는 어려움이 있다. 이를 해결하기 위해..

논문 제목 : Learning deep features for discriminative localization 논문 주소 : openaccess.thecvf.com/content_iccv_2017/html/Selvaraju_Grad-CAM_Visual_Explanations_ICCV_2017_paper.html ICCV 2017 Open Access Repository Grad-CAM: Visual Explanations From Deep Networks via Gradient-Based Localization Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, Dhruv Batra; Pr..

논문 제목 : Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization 논문 주소 : openaccess.thecvf.com/content_iccv_2017/html/Selvaraju_Grad-CAM_Visual_Explanations_ICCV_2017_paper.html ICCV 2017 Open Access Repository Grad-CAM: Visual Explanations From Deep Networks via Gradient-Based Localization Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, De..