반응형

 

 

1. D2-Net (https://arxiv.org/pdf/1905.03561.pdf)

: CNN 기반

 

 

2. LF-Net (https://papers.nips.cc/paper/7861-lf-net-learning-local-features-from-images.pdf)

Local feature를 학습

3. Feature matching CNN 기반 (https://ieeexplore.ieee.org/abstract/document/8780936)

 

 

4. Deep Grapthical feature learning for feature matching (https://openaccess.thecvf.com/content_ICCV_2019/papers/Zhang_Deep_Graphical_Feature_Learning_for_the_Feature_Matching_Problem_ICCV_2019_paper.pdf)

GNN 기반 feature point를 local feature 로 변환. 이로 인해 feature matching 시에 inference 알고리즘이 단순화됨.

 

 

 

 

출처 : https://towardsdatascience.com/image-feature-extraction-traditional-and-deep-learning-techniques-ccc059195d04

반응형

'Programming > SLAM' 카테고리의 다른 글

iMap : Implicit Mapping and Positioning in Real-Time 논문 리뷰  (0) 2023.05.21
DROID SLAM 논문 리뷰  (0) 2023.05.20
SLAM 비선형 최적화 기법  (0) 2023.05.01
NICE-SLAM 논문리뷰  (0) 2023.03.28
Optical Flow란?  (0) 2023.02.19

+ Recent posts