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 | 31 |
Tags
- Markov transition matrix
- learning to prompt
- ENERGY-BASED MODELS FOR CONTINUAL LEARNING
- prompt learning
- PnP algorithm
- requires_grad
- learning to prompt for continual learning
- CVPR2022
- 베이지안 정리
- L2P
- Facial Landmark Localization
- timm
- state_dict()
- DualPrompt
- VQ-diffusion
- Face Pose Estimation
- Face Alignment
- Continual Learning
- Mask diffusion
- Mask-and-replace diffusion strategy
- Energy-based model
- img2pose: Face Alignment and Detection via 6DoF
- CIL
- Vector Quantized Diffusion Model for Text-to-Image Synthesis
- VQ-VAE
- mmcv
- Img2pose
- Discrete diffusion
- Class Incremental
- Class Incremental Learning
Archives
- Today
- Total
Computer Vision , AI
[One-page summary] Make A Video: Text to Video Generation Without Text Video Data (ICLR 2023) by Singer et al. 본문
Paper_review[short]
[One-page summary] Make A Video: Text to Video Generation Without Text Video Data (ICLR 2023) by Singer et al.
Elune001 2024. 1. 16. 01:02● Summary: Text to Video generation with Text Image Data
● Approach highlight
- Text-to-Image Model: DALLE 2 architecture
- Spatiotemporal layers: U-Net based spatiotemporal diffusion decoder makes a frame from noise
- Frame interpolation network
● Main Results:
● Discussion
- How to generate temporal frames from the spatiotemporal decoder
- How to learn the relationship between text and action that can only be inferred in videos (ex. a video of a person waving their hand left to right or right to left)
'Paper_review[short]' 카테고리의 다른 글
'Paper_review[short]' Related Articles
more
