Paper Review
-
Convolutional Neural Network
-
Natural Language Processing, 자연어처리
- Attention Is All You Need : Transformer
-
3D/Visual Geometry, 영상 기하학
- Structure from Motion
-
그래프 뉴럴 네트워크 (Graph Neural Network)
- 그래프 뉴럴 네트워크 (Graph Neural Network)
- Graph Attention Networks
- Graph Convolutional Networks
- Original Paper / Summary PDF / Video Presentation / Code Practice
-
Recommender System, 추천 시스템
- Recommender System
-
Visual Localization, 카메라 포즈 위치 추정
- Visual Localization 소개
- Introduction to Camera Pose Estimation with Deep Learning
- Summary PDF / Paper review video
- Introduction to Camera Pose Estimation with Deep Learning
- Visual Navigation
- Neural Topological SLAM for Visual Navigation
- Summary PDF / Paper review video
- Neural Topological SLAM for Visual Navigation
- 2015,2016,2017년도 End-to-End Visual Localization
- Baysian PoseNet, LSTM-Pose, VidLoc, Hourglass-Pose
- Summary PDF / Paper review video
- Baysian PoseNet, LSTM-Pose, VidLoc, Hourglass-Pose
- 2017, 2019년도 End-to-End Visual Localization
- PoseNet2, AtLoc
- Summary PDF / [Paper review video]
- PoseNet2, AtLoc
- 2021년도 End-to-End Visual Localization
- CaTiLoc
- Original Paper Link / Summary PDF / Paper review video
- CaTiLoc
- Hierarchical Structure
- 카메라 위치 파악을 위한 복셀 표현의 covisibility 기반 참조 이미지 클러스터링(2022)
- Original Paper Link / Summary PDF / Summary PDF / Paper review video 1 / Paper review video2
- Efficient Large-scale Localization by Global Instance Recognition(2022)
- Original Paper Link / Summary PDF/ Paper review video
- From Coarse to Fine: Robust Hierarchical Localization at Large Scale(2019)
- Original Paper Link / Summary PDF / Paper review video
- 카메라 위치 파악을 위한 복셀 표현의 covisibility 기반 참조 이미지 클러스터링(2022)
- Visual Localization 소개
-
Visual Place Recognition
- NetVLAD : CNN architecture for weakly supervised place recognition
- Original Paper Link / Summary PDF/ Paper review video
- Patch-NetVLAD: Multi-Scale Fusion of Locally-Global Descriptors for Place Recognition
- Original Paper Link / Summary PDF/ Paper review video
- NetVLAD : CNN architecture for weakly supervised place recognition
-
Contrastive Learning
-
Unsupervised / Selsupervised Learning
- A Simple Framework for Contrastive Learning of Visual Representation
- Momentum Contrast for Unsupervised Visual Representation Learning
- Prototypical Contrastive Learning of Unsupervised Representations
-
Supervised Learning
- Supervised Contrastive Learning
-
Novel View Synthesis
- NeRF(2020)
- 2D 이미지를 인공지능써서 3D로 만드는 방법
- Instant NeRF(2022)