일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | |||
5 | 6 | 7 | 8 | 9 | 10 | 11 |
12 | 13 | 14 | 15 | 16 | 17 | 18 |
19 | 20 | 21 | 22 | 23 | 24 | 25 |
26 | 27 | 28 | 29 | 30 | 31 |
- Artificial Intelligence
- 딥러닝
- Class activation map
- 코딩 테스트
- meta-learning
- Deep learning
- xai
- Machine Learning
- 기계학습
- 설명가능한
- SmoothGrad
- 백준
- coding test
- Cam
- python
- 설명가능한 인공지능
- AI
- grad-cam
- keras
- 코딩테스트
- GAN
- Interpretability
- Unsupervised learning
- 인공지능
- cs231n
- 머신러닝
- 메타러닝
- Explainable AI
- 시계열 분석
- Score-CAM
- Today
- Total
iMTE
논문 제목 : Sanity checks for saliency maps 논문 주소 : arxiv.org/abs/1810.03292 Sanity Checks for Saliency Maps Saliency methods have emerged as a popular tool to highlight features in an input deemed relevant for the prediction of a learned model. Several saliency methods have been proposed, often guided by visual appeal on image data. In this work, we propose an a arxiv.org 주요 내용 : 1) Saliency map은 학..
논문 제목 : Sanity checks for saliency maps 논문 주소 : arxiv.org/abs/1810.03292 Sanity Checks for Saliency Maps Saliency methods have emerged as a popular tool to highlight features in an input deemed relevant for the prediction of a learned model. Several saliency methods have been proposed, often guided by visual appeal on image data. In this work, we propose an a arxiv.org 주요 수식 정리: 0) Definition in..
논문 제목 : SmoothGrad : removing noise by adding noise 논문 주소 : arxiv.org/abs/1706.03825 SmoothGrad: removing noise by adding noise Explaining the output of a deep network remains a challenge. In the case of an image classifier, one type of explanation is to identify pixels that strongly influence the final decision. A starting point for this strategy is the gradient of the class score arxiv.org 주요 ..
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..