-
## 一言でいうと
物理的な敵対的サンプルを作成する手法。カメラのレンズに特殊な柄のステッカーを貼ることで、DNNベースの物体認識モデルの判断を誤らせることが可能。
### 論文リンク
https://arxiv.org/abs/1904.00759
### 著者/所属機関
Juncheng Li, Frank R. Schmidt, J. Zico Kolter
* Bo…
-
Organizing papers according to their contents is good, but chronological order makes it easier to follow the new trend in top conferences. We can put several markdown files in this repo, each named as…
-
## 論文リンク
https://arxiv.org/abs/1901.09192
## 公開日（yyyy/mm/dd）
2019/01/26
ICML 2019
## 概要
Rejection option付きのDNN classifier(SelectiveNet)の提案。Classifierとrejectionの両方をend-to-endで学習する。（既存手法は事前学習済…
-
https://github.com/Tiiiger/SGC official implementation for the paper "Simplifying Graph Convolutional Networks" on ICML 2019 by Wu, Felix and Souza, Amauri and Zhang, Tianyi and Fifty, Christopher and…
-
- [ ] 论文阅读情况
- ICML 2019 MASS: Masked Sequence to Sequence Pre-training for Language Generation
> https://arxiv.org/pdf/1905.02450.pdf
-
Hello, thanks for you nice paper and released source code.
I wonder how did you select K-shot samples in Pascal VOC dataset for novel classes.
Is it same with FSRW(ICCV 2019) and TFA(ICML 2020) ..?
-
Hi Jenny,
I find it really interesting of this GNF work. And I have a question in the model design for the graph generation in Figure 2 of the original NeurIPS 2019 paper.
, and want to implement the DeepFactor model using GluonTS package. And, I have a question about How to assume a…