ZZIN33
re-code-cord
ZZIN33
전체 방문자
오늘
어제
  • 분류 전체보기 (52)
    • Paper (4)
      • Generative Model (2)
      • Segmentation (1)
      • 모델 경량화 (1)
    • Study (34)
      • AI (10)
      • MLOps (8)
      • CS (4)
      • OpenCV (1)
      • Algorithm (9)
      • ETC (2)
    • Project (6)
    • ETC (8)
      • 부스트캠프 AI Tech (2)
      • 도서 리뷰 (5)

블로그 메뉴

  • Home
  • About
  • Github

인기 글

최근 댓글

최근 글

티스토리

hELLO · Designed By 정상우.
ZZIN33

re-code-cord

Study/AI

Deep Learning Library for video understanding

2021. 11. 30. 15:04

Representative Library for video understanding

  • facebookresearch/PyTorchVideo [github]
  • open-mmlab/MMAction2 [github]
  • paddlepaddle/PaddleVideo [github]

 


Compare Libraries

 

PyTorchVideo

MMAction2

PaddleVideo

Support
Model

- Slow & SlowFast
- X3D 
- MViT
- Slow & SlowFast
- X3D

- TimeSformer
- BMN
- PoseC3D
- PP-TSM
- AGCN
- TimesFormer
- BMN
- AttentianLSTM
Feature - Support Accelerator for optimizing target device
- Using object detector from Detectron2
- Support 4 video understanding tasks 
- Modular design
 - Support 3 video understanding tasks
- Support Faster training stratege

 

저작자표시 (새창열림)

'Study > AI' 카테고리의 다른 글

데이터가 충분하다고 말하려면 얼마나 있어야 할까?  (0) 2022.01.24
딥러닝이란 무엇일까?  (0) 2022.01.15
Knowledge Distillation 구현  (0) 2021.11.29
[Pytorch] Tips for Loading Pre-trained Model  (0) 2021.11.27
Lightweight Deep Learning  (0) 2021.11.24

    티스토리툴바