본문 바로가기 메뉴 바로가기

note

프로필사진
  • 글쓰기
  • 관리
  • 태그
  • 방명록
  • RSS

note

검색하기 폼
  • 분류 전체보기 (38)
    • Deep Learning (14)
    • digital signal process (3)
    • MLOps (6)
    • OS (5)
    • algorithm (2)
    • error (0)
    • etc (8)
    • DAP (0)
      • SQLD (0)
  • 방명록

camera geometry (1)
Unsupervised learning of depth and ego-motion from video

이 논문은 unsupervised한 방법으로 image depth를 구하는 모델을 학습시키는 방법이다. 학습에는 sequential한 이미지가 필요하고, 각 이미지의 pose를 이용해 한 이미지의 point들을 다른 이미지로 projection하고, projection된 위치의 점과 Intensity를 비교해 loss를 구한다. 여기서 pose도 model output이므로 gt없이 loss 계산이 가능해진다. Overview 조금 더 구체적으로 살펴보면 다음과 같은 단계로 나누어 생각해 볼 수 있다. 1. $t$에서의 이미지 $I_{t}$를 depth CNN에 통과시켜 Depthmap $\hat{D_{t}}(p)$를 구한다.(Depth map은 모든 이미지 픽셀에 대한 depth를 가지고 있다.) 2...

Deep Learning 2022. 2. 17. 00:22
이전 1 다음
이전 다음
최근에 올라온 글
TAG
  • depth
  • deeplearning
  • image
  • 3d object detection
  • conditional GAN
  • depthmap
  • 신호처리
  • Generative Model
  • 운영체제
  • machine learning
  • DSP
  • Gan
  • feature
  • Raspberry Pi
  • Depth estimation
  • TRACKING
  • 디지털신호처리
  • 딥러닝
  • controllable GAN
  • OS
  • AI
  • Building Basic GAN
  • Operating System
  • ML
  • MLOps
  • design pattern
  • ML Pipeline
  • Deep learning
  • mode collapse
  • pcb
more
Total
Today
Yesterday
최근에 달린 댓글
링크
공지사항
«   2025/05   »
일 월 화 수 목 금 토
1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31
글 보관함

Blog is powered by Tistory / Designed by Tistory

티스토리툴바