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 |
Tags
- ddpm
- Few-shot generation
- UE5
- deep learning
- RNN
- GAN
- WinAPI
- 모션매칭
- NLP
- Stat110
- Unreal Engine
- 생성모델
- Diffusion
- WBP
- 폰트생성
- 언리얼엔진
- 오블완
- motion matching
- Font Generation
- animation retargeting
- ue5.4
- Generative Model
- dl
- userwidget
- 딥러닝
- CNN
- BERT
- multimodal
- 디퓨전모델
- cv
Archives
- Today
- Total
목록Neural Style Transfer (1)
Deeper Learning

Neural Style Transfer 이미지 C와 content가 비슷하고 이미지 S와 style이 비슷한 이미지 G를 만드는 알고리즘 Cost function의 정의는 다음과 같다. hyperparameter α와 β에 의해 content loss, style loss에 weight가 적용된 Cost function Content Cost Function a(C) = input C가 hidden layer를 통과한 activations (1, n_h, n_w, n_c) a(G) = input G가 hidden layer를 통과한 activations (1, n_h, n_w, n_c) (n_h,n_w,n_c) = hidden layer의 output shape ..
AI/Deep Learning
2020. 11. 28. 17:15