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
- learning to prompt
- Img2pose
- Discrete diffusion
- Markov transition matrix
- ENERGY-BASED MODELS FOR CONTINUAL LEARNING
- VQ-diffusion
- Class Incremental Learning
- L2P
- Class Incremental
- timm
- 베이지안 정리
- Vector Quantized Diffusion Model for Text-to-Image Synthesis
- VQ-VAE
- Face Pose Estimation
- Facial Landmark Localization
- mmcv
- Continual Learning
- prompt learning
- DualPrompt
- Mask diffusion
- state_dict()
- requires_grad
- learning to prompt for continual learning
- img2pose: Face Alignment and Detection via 6DoF
- CVPR2022
- Mask-and-replace diffusion strategy
- Energy-based model
- PnP algorithm
- CIL
- Face Alignment
Archives
- Today
- Total
Computer Vision , AI
Paint by Example: Exemplar-based Image Editing with Diffusion Models (CVPR 2023) by Yang et al. 본문
Paper_review[short]
Paint by Example: Exemplar-based Image Editing with Diffusion Models (CVPR 2023) by Yang et al.
Elune001 2024. 1. 16. 01:37● Summary: Simple method for Image editing with a diffusion model only using CLIP [CLS] token embedding
● Approach highlight
- Image editing without labels using only the detection model
- Crop the original image and augment the image for CLIP embedding
- Only use [CLS] token to prevent the model from just doing copy-and-paste
- Classifier free sampling for image identity (scale factor)
● Main Results
● Discussion
- Could it have been possible to use the entire CLIP embedding instead of just the [CLS] token, but in a reduced form?
'Paper_review[short]' 카테고리의 다른 글
'Paper_review[short]' Related Articles
more
