Notice
Recent Posts
Recent Comments
Link
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
- novel view synthesis
- instructany2pix
- 네이버 부스트캠프 ai tech 6기
- text-to-video diffusion
- magic clothing
- 프로그래머스
- sound-to-image generation
- BOJ
- dreamfusion
- sonicdiffusion
- 3d generation
- Programmers
- DP
- 코딩테스트
- VirtualTryON
- 3d editing
- Vit
- transformer
- diffusion
- 3d gaussian splatting
- dreamgaussian
- 논문리뷰
- posterior distillation sampling
- visiontransformer
- syncnoise
- text2room
- 코테
- insturctnerf2nerf
- 3d editting
- Python
Archives
- Today
- Total
목록2024/04/10 (1)
평범한 필기장
![](http://i1.daumcdn.net/thumb/C150x150/?fname=https://blog.kakaocdn.net/dn/b1RWeA/btsGrbuinJh/xn75khV0tKklTNh1t5VMY1/img.png)
1. Introduction 기존 diffusion models는 LSUN과 ImageNet과 같은 어려운 generation에서는 GAN (BIgGAN-deep)에 경쟁이 되지 않는 FID score를 냈다. 본 논문에서 diffusion models와 GANs사이의 차이는 (1) 최신 GAN의 architecture는 고도로 연구되고 refine되었다는 것과 (2) GANs는 다양성을 fidelity로 맞바꿀 수 있다는 것이다. 본 논문에서는 이 두 가지의 이점을 가져오는 것을 목표로 한다. (1)은 모델 아키텍쳐를 향상시킴으로써 (2)는 다양성을 fidelity로 맞바꾸는 계획을 구상함으로써 해결하려한다. 이를 통해 몇 개의 metric과 dataset에서 GAN을 뛰어넘는 sota를 달성했다고 한..
AI/Diffusion
2024. 4. 10. 00:26