티스토리
관련 서비스
  • Symbols스토리 홈바로가기바로가기
  • 브런치브런치스토리바로가기바로가기
  • 카카오스토리카카오스토리바로가기바로가기

딥러닝 공부방
검색하기
블로그 홈

블로그 홈

딥러닝 공부방

deep-learning-study.tistory.com/m
신고

까먹으면 다시 보려고 정리하는 블로그

구독자
406
방명록 방문하기
구독하기
반응형

주요 글 목록

  • [논문 읽기] CaSP, Class agnostic Semi-Supervised Pretraining for Detection and Segmentation CaSP, Class agnostic Semi-Supervised Pretraining for Detection and Segmentation https://arxiv.org/abs/2112.04966 CaSP: Class-agnostic Semi-Supervised Pretraining for Detection and Segmentation To improve instance-level detection/segmentation performance, existing self-supervised and semi-supervised methods extract either very task-unrelated or very task-specific training signals from unlabeled d.. 공감수 1 댓글수 0 2022. 3. 2.
  • [논문 읽기] PseudoSeg, Designing Pseudo Labels for Semantic Segmentation(2020) PseudoSeg, Designing Pseudo Labels for Semantic Segmentation(2020) semi segmentation 논문. pixel label이 존재하면, 이미지에 weak augmentation을 준 뒤 모델로 전달하여 얻은 prediction과 gt사이의 cross entropy loss를 계산. unlabeled image에 대해서는 strong augmentation을 가한뒤에 얻은 prediction과 pseudo label 사이의 cross entropy를 계산한다. 그러면 pseudo label을 어떻게 얻을까? grad-CAM과 decode의 출력값을 활용한다. grad-CAM은 prediction에 높은 영향력이 있는 region을 검출하는데, 이는 .. 공감수 1 댓글수 0 2022. 2. 20.
  • [논문 읽기] Soft Teacher(2021), End-to-End Semi-Supervised Object Detection with Soft Teacher End-to-End Semi-Supervised Object Detection with Soft Teacher https://arxiv.org/abs/2106.09018 End-to-End Semi-Supervised Object Detection with Soft Teacher This paper presents an end-to-end semi-supervised object detection approach, in contrast to previous more complex multi-stage methods. The end-to-end training gradually improves pseudo label qualities during the curriculum, and the more and .. 공감수 0 댓글수 0 2022. 1. 30.
  • [논문 읽기] Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision(2021) Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision https://arxiv.org/abs/2106.01226 Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision In this paper, we study the semi-supervised semantic segmentation problem via exploring both labeled data and extra unlabeled data. We propose a novel consistency regularization approach, called cross pseudo supervision (CPS). Ou.. 공감수 1 댓글수 0 2022. 1. 29.
  • [논문 읽기] ReCo(2021), Bootstrapping Semantic Segmentation with Regional Contrast Bootstrapping Semantic Segmentation with Regional Contrast PDF, Segmentation Contrastive learning in semi supervision, Shikun, et al, arXiv 2021 Summary 해당 논문은 segmentation에 contrastive learning을 적용하여 supervised or semi learning을 수행합니다. contrative learning은 positive 와 negative를 정의해주고 loss를 설계하는데 segmentation에서는 어떻게 positive와 negative를 정의하는지 확인해보면 좋을 듯 싶네요 또 봐야할 점은 image 내의 모든 pixel에 대하여 similari.. 공감수 1 댓글수 0 2021. 9. 29.
  • [Paper review] Mean teachers are better role models(2017) Mean teachers are better role models: Weight-averaged consistency targets imporve semi-supervised deep learning results Antti Tarvainen, Harri Valpola, arxiv 2017 PDF, Semi Supervised Learning By SeonghoonYu July 18th, 2021 Summary Previous best performance model of semi-supervised learning is Temporal Ensembling having a problem. Since each target is updated only once per epoch, the learned inf.. 공감수 2 댓글수 0 2021. 7. 18.
  • [Paper review] Temporal Ensembling for Semi-Supervised Learning(2016) Temporal Ensembling for Semi-Supervised Learning Samuli Laine, Timo Aila, arxiv 2016 PDF, Semi-Supervised Learning By SeonghoonYu July 18th, 2021 Summary They propose $\sqcap$-model and temporal Ensemling in a semi-supervised learning setting only a small portion of training data is labeled. During training, $\sqcap$-model evaluates each training input $x_i$, resulting in prediction vetors $z_i$.. 공감수 2 댓글수 0 2021. 7. 18.
  • [논문 읽기] (2019), Consistency-based Semi-supervised Learning for Object Detection 안녕하세요, 오늘 읽은 논문은 Consistency-based Semi-supervised Learning for Object Detection 입니다. object detection task는 많은 수의 annotated sample이 필요합니다. 그리고 이를 사람이 직접 annotate하는 데에는 많은 비용이 필요합니다. 해당 논문에서는 unlabeled data를 활용하기 위한 semi-supervised learning 방법을 제안합니다. unlabeled data를 활용하기 위해 (1) consistency loss를 제안하고, back-ground class가 성능에 악영향을 주는 것을 방지하기 위해 (2) Background Elimination(BE)를 제안합니다. labeleing cos.. 공감수 2 댓글수 0 2021. 7. 8.
    반응형
    문의안내
    • 티스토리
    • 로그인
    • 고객센터

    티스토리는 카카오에서 사랑을 담아 만듭니다.

    © Kakao Corp.