- 목적: CS231n 강의에서 다룬 주요 논문 31개(CNN basics, Computer Vision basics, RNN basics)를 함께 읽고 리뷰합니다.
- 기간: 미정
- 참여자: 현재 모집 중입니다.
- 기획 그룹: AI Robotics KR
- 온라인(구글 행아웃)으로 매주 ㅇ요일 ㅇ시 ~ ㅇ시에 진행됩니다.
- 논문 하나를 읽고 스터디전날까지 5줄 요약과 질문 1개를 제출하고 스터디날 다같이 질문에 대한 답을 찾습니다.
- 논문이 여러개 묶인 날은 하나씩 발표자를 정해 리뷰하고 다른 스터디원들은 그중 논문 하나만 읽고 5줄 요약, 질문 1개를 스터디전날까지 제출합니다.
- 팀원들은 Slack 채널(study_cs231n_papers)과 카카오톡 그룹챗을 이용하여 소통합니다.
(현재 스터디 계획은 초안이라 구체적인 스터디 계획은 스터디원분들과 상의후 변경)
- ['Paper Review Records' 폴더]: 스터디원들의 논문 리뷰(5줄 요약)를 공유하는 공간입니다.
- [Issues]: 스터디에서 나눈 질의응답이나 추가적인 정보를 공유하는 공간입니다.
- ImageNet Classification with Deep Convolutional Neural Networks(2012) [AlexNet]
- Very deep convolutional networks for large-scale image recognition (2014)[VGGNet]
- Going deeper with convolutions (2014)[GoogLeNet]
- Visualizing and understanding convolutional networks(2013)[ZFnet]
- Network in Network (2014)
- Delving Deep into Rectifiers: surpassing human-level performance on ImageNet classification (2015) / PR020 video
- Rethinking the Inception Architecture for Computer vision
- Batch Normalization (2015) / PR-021 video
- Deep residual learning for image recognition (2015) [ResNet] / PR-170 video
- Identity mappings in Deep residual networks (2016)
- Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning (2016)
- Densely Connected Convolutional Networks (2017) / PR-028 video
- MobileNets: efficient convolutional neural networks for mobile vision applications (2017) / PR-044 video
- DeCAF: A deep convolutional activation feature for generic visual recognition (2013)
- Rich feature hierarchies for accurate object detection and semantic segmentation
- Fully Convolutional Networks for Semantic Segmentation
- Learning Deconvolution Network for Semantic Segmentation
- Fast R-CNN (2015)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks / PR-012 video
- SSD: Single-shot multibox detector / PR-132 video
- You Only Look Once / PR-016 video
- mask R-CNN / PR-057 video
- Focal loss for dense object detection
- Long Short Term Memory
- Sequence to Sequence learning with Neural networks (2014)
- Learning phrase representations using rnn encoder-decoder for statistical machine translation / PR-003 video
- Long-term recurrent convolutional networks for visual recognition and description
- Recurrent models of visual attention
- show , attend and tell: Neural Image caption generation with visual attention
- show and tell: a neural image caption generator / PR-041 video
- Attention is all you need / PR-049 video
