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
- Facial Landmark Localization
- L2P
- Mask diffusion
- Continual Learning
- Vector Quantized Diffusion Model for Text-to-Image Synthesis
- VQ-VAE
- CVPR2022
- Face Alignment
- Mask-and-replace diffusion strategy
- Discrete diffusion
- learning to prompt
- VQ-diffusion
- PnP algorithm
- Face Pose Estimation
- timm
- requires_grad
- prompt learning
- Img2pose
- Energy-based model
- state_dict()
- Class Incremental Learning
- Class Incremental
- 베이지안 정리
- Markov transition matrix
- learning to prompt for continual learning
- DualPrompt
- mmcv
- img2pose: Face Alignment and Detection via 6DoF
- CIL
- ENERGY-BASED MODELS FOR CONTINUAL LEARNING
Archives
- Today
- Total
Computer Vision , AI
[One-page summary] From Motor Control to Team Play in Simulated Humanoid Football (Science Robotic 2022) by Liu et al. 본문
Paper_review[short]
[One-page summary] From Motor Control to Team Play in Simulated Humanoid Football (Science Robotic 2022) by Liu et al.
Elune001 2024. 1. 16. 01:24● Summary: hierarchically structured behavior and long-horizon coordination for RL
● Approach highlight
- Hierarchically structured behavior
- Imitation for low level ex)run
- Reinforcement learning for Drill ex) kick, dribble
- Distillation for single player
- Multi player reinforcement learning
● Main Results:
● Discussion
- Limitation of simple reward only goal score)
- Too heavy model