Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[20220501] Weekly AI ArXiv 만담 #50

Closed
jungwoo-ha opened this issue Apr 30, 2022 · 4 comments
Closed

[20220501] Weekly AI ArXiv 만담 #50

jungwoo-ha opened this issue Apr 30, 2022 · 4 comments

Comments

@jungwoo-ha
Copy link
Owner

jungwoo-ha commented Apr 30, 2022

News

Arxiv

@ghlee0304
Copy link

ghlee0304 commented Apr 30, 2022

Arxiv

@veritas9872
Copy link

veritas9872 commented May 1, 2022

OpenGlue: Open Source Graph Neural Net Based Pipeline for Image Matching

image

Kaggle: https://www.kaggle.com/datasets/ksork6s4/openglue
Arxiv: https://arxiv.org/abs/2204.08870
GitHub: /~https://github.com/ucuapps/OpenGlue

Understanding The Robustness in Vision Transformers

image

Screenshot (50)

Arxiv: https://arxiv.org/abs/2204.12451
GitHub: /~https://github.com/NVlabs/FAN

@hollobit
Copy link

hollobit commented May 1, 2022

ISO/IEC JTC 1/SC 42(Artificial Intelligence) plenary week (4/18 ~ 4/30)

https://www.iso.org/committee/6794475.html

  • 참여자 240명, 11개 표준 제정, 26개 표준 개발 중
    • 3개 timezone rotation : 아침6시, 오후14시, 오후22시
    • WG1: Foundational. WG2: Data, WG3 : Trustworthiness, WG4: Application, WG5: Computational Approach
  • 선도그룹: 미국, 중국, 일본, 영국
    • JWG 2 활동 시작: Testing of AI-based system (영국, 한국)
  • 추격그룹: 인도, 프랑스, UAE
    • UAE: "Standards for Unsupervised Learning" - unsupervised + RL + decision transformer"
    • UAE: "Standards for Explainability in AI"
    • 인도 : verification and validation analysis of AI systems"
    • 프랑스 : AG3
  • 배후세력

Google FormNet

https://ai.googleblog.com/2022/04/formnet-beyond-sequential-modeling-for.html

@jwlee-ml
Copy link

jwlee-ml commented May 1, 2022

DeiT III: Revenge of ViT (밀린 숙제)

  • https://arxiv.org/abs/2204.07118
  • ViT 계의 ResNet strikes back 같은 논문
  • 흥행에는 실패한 것 같기도?
  • 몇가지 Training Technique을 활용해서 vanilla ViT 성능을 끌어올린(쥐어짠) 논문
  • 사용한 방법들
    Stochastic Depth, LayerScale, Binary Cross Entropy, Data Augmentation, Croping

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants