johyeongseob 님의 블로그

[논문 리뷰] SigLIP: Sigmoid Loss for Language Image Pre-Training

논문: https://arxiv.org/abs/2303.15343저자: Xiaohua Zhai$^*$, Basil Mustafa, Alexander Kolesnikov, Lucas Beyer$^*$$^*$Equal contribution, Google DeepMind, Z¨urich, Switzerland, {fxzhai, basilm, akolesnikov, lbeyerg}@google.com인용: Zhai, Xiaohua, et al. "Sigmoid loss for language image pre-training." Proceedings of the IEEE/CVF international conference on computer vision. 2023.깃허브: https://github.com/..

[논문 리뷰] Deep One-Class Classification

논문: https://proceedings.mlr.press/v80/ruff18a.html저자: Lukas Ruff* $^1$, Robert A. Vandermeulen* $^2$, Nico Gornitz $^3$, Lucas Deecke $^4$, Shoaib A. Siddiqui $^{2, 5}$, Alexander Binder $^6$, Emmanuel Muller $^1$, Marius Kloft $^2$A part of the work was done while LR, RV, LD, and MK were with Department of Computer Science, Humboldt University of Berlin, Germany. $^1$ Hasso Plattner Institute, ..

논문: https://arxiv.org/abs/2305.06500저자: Wenlinag $^{†,1,2,*}$ , Junnan Li $^{†,1,*}$ , Dongxu Li $^{1}$ , Anthony Meng Huat Tiong $^{1,3}$ , Junqi Zhao $^{3}$ , Weisheng $^{3}$ , Boyang Li $^{3}$ , Pascale Fung $^{2}$ , Steven Hoi $^{1}$ $^{1}$ Salesforce Research, $^{2}$ Hong Kong University of Science and Technology, $^{3}$ Nanyang Technological University, Singapore$^{†}$ Equal contribut..

논문: https://arxiv.org/abs/2301.12597저자: Junnan Li, Dongxu Li, Silvio Savarese, Steven Hoi.Salesforce Research인용: Li, Junnan, et al. "Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models." International conference on machine learning. PMLR, 2023.코드: https://github.com/salesforce/LAVIS/tree/main/projects/blip2 0. 초록Vision-and-language pre-training ..

논문: https://arxiv.org/abs/2201.12086저자: Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi.Salesforce Research인용: Li, Junnan, et al. "Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation." International conference on machine learning. PMLR, 2022.코드: http://github.com/salesforce/BLIP 0. 초록Vision-Language Pre-training (VLP) 는 이미지-언어 작업 대부분에 사용된다. 그러..

논문: https://arxiv.org/abs/1711.00937저자: Aaron van den Oord, Oriol Vinyals, Koray Kavukcuoglu. DeepMind, {avdnoord, vinyals, korayk} @google.com인용: Van Den Oord, Aaron, and Oriol Vinyals. "Neural discrete representation learning." Advances in neural information processing systems 30 (2017). 0. 초록지도학습 (supervision) 없이 이미지로부터 유용한 표현을 학습하는건 어렵다. 이번 논문에서 저자들은 이산 표현을 학습하여 간단하지만 강력한 생성형 모델을 제안한다. Vecto..

📌 이 글에 대하여 이 게시글은 Kaggle의 Moein Shariatnia 님이 작성한 원문 글을 한국어로 번역한 것입니다.원문은 Apache License 2.0 하에 공개되었으며, 이 블로그 역시 해당 라이선스를 따릅니다.원문 저자:Moein Shariatnia원문 위치: Kaggle Notebook라이선스: Apache License 2.0 전문 보기본 번역은 비상업적/교육적 목적이며, 원문 저자의 저작권과 라이선스를 존중합니다.코드: https://github.com/johyeongseob/from-scratch-ai데이터셋: https://www.kaggle.com/datasets/hsankesara/flickr-image-dataset라이브러리 설치conda create -n clip-en..

📌 이 글에 대하여이 게시글은 Kaggle의 Sushant Kumar 님이 작성한 원문 글을 한국어로 번역한 것입니다.원문은 Apache License 2.0 하에 공개되었으며, 이 블로그 역시 해당 라이선스를 따릅니다.원문 저자: Sushant Kumar 원문 위치: Kaggle Notebook 라이선스: Apache License 2.0 전문 보기본 번역은 비상업적/교육적 목적이며, 원문 저자의 저작권과 라이선스를 존중합니다.코드: https://github.com/johyeongseob/from-scratch-aidependency conflict 확인Windows11, Python 3.8.18, torch version: 2.4.1+cu121, CUDA: 12.1, GPU: NVIDIA GeFo..

논문: https://arxiv.org/pdf/2010.11929저자: Alexey Dosovitskiy∗$^†$, Lucas Beyer∗, Alexander Kolesnikov∗, Dirk Weissenborn∗, Xiaohua Zhai∗, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby∗$^†$∗ equal technical contribution, $^†$ equal advising, Google Research, Brain Team인용: Dosovitskiy, Alexey, et al. "An image is worth 16x16 word..

논문: https://arxiv.org/abs/1706.03762저자: Ashish Vaswani* (Google Brain), Noam Shazeer∗ (Google Brain), Niki Parmar∗(Google Research), Jakob Uszkoreit∗ (Google Research), Llion Jones∗ (Google Research), Aidan N. Gomez∗$^†$ (University of Toronto), Łukasz Kaiser∗ (Google Brain), Illia Polosukhin∗$^‡$∗ Equal contribution, $^†$ Work performed while at Google Brain. $^‡$ Work performed while at Google..