Transfer Learning
https://www.youtube.com/watch?v=yofjFQddwHE -Transfer learning (전이학습) x,y쌍을 학습시킨다 x는 이미지, y는 라벨 신경망의 마지막 출력층을 삭제하고 마지막 레이어에 대해 무작위로 초기화된 세트를 만든다. ex)x-ray결과같은거 x방사선이미지 y예측질병 , 마지막층의 가중치와 바이어스 를 재훈련시킨다. 앞부분의 줄기는 그대로 가져감 Rule of Thumb ; 경험의 법칙 ;과학적사실이 아닌 경험에 바탕한 결정이란 뜻 (기존에 학습된 부분을 비유하는 것 같음) 이미지인식에서의 첫번재 훈련단계는 pretrained 라고 함 이것을 모두 갱신한 뒤 방사선데이터에 대해 훈련하는 세부조정을 fine tuning이라고 함 학습레이어의 앞부분은 로우레벨로 인..
AI/주워들은 것들
2021. 11. 7. 22:46
공지사항
최근에 올라온 글
최근에 달린 댓글
- Total
- Today
- Yesterday
링크
TAG
- alternatives.log
- tf version porting
- TransferLearning
- Group Acitivty Recognition
- 샴네트워크
- max-margin_loss
- /var/log/faillog
- neural network applications
- Chatgpt4
- 도메인 어댑테이션
- 전이학습
- tf2to1
- batch란
- 1-stage detection
- torchmodel
- Source-free unsupervised domain adaptation
- Arbitrary image stylization
- ctfloader
- /var/log/alternatives.log
- 딥페이크탐지
- yolov3 #tensorflow_convert
- StyleCLIP
- DetectoRS
- /var/log/tallylog
- textual guidance
- Deformable Convolutional Networks
- Deformable Part Model
- 첫논문
- tensorflow convert
- /var/log/fontconfig.log
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 |
글 보관함