Alibaba-MIIL / STAM

Official implementation of "An Image is Worth 16x16 Words, What is a Video Worth?" (2021 paper)

Alibaba-MIIL/STAM Issues

Could you please share training hyper-parameters?
Updated a year ago1
train model
Closed a year ago
How to train?
Updated a year ago
about temporal module
Closed a year ago
39.8% of the validation data is not used for performance test
Closed 3 years ago2
how to prepare a validation folder?
Closed 2 years ago1
python -m infer 预测是上传任何视频都行吗
Updated 3 years ago1
Performance with low hyper parameters
Updated 3 years ago
Why did you use a pytorch built-in TransformerEncoder in TAggregate module?
Updated 3 years ago
Pretrain weights from the ImageNet
Closed 3 years ago2
时间聚合时维度如何对齐？
Closed 3 years ago2
Linear Projection
Closed 3 years ago2
when will you update? looking forward to your code !
Closed 3 years ago
Cannot access pretrained model link
Closed 3 years ago1
是把16帧组成当成一张图片吗，另外预训练模型会公布吗
Updated 3 years ago
some training hyperparameters about kinetics400
Closed 3 years ago1
Training code
Closed 3 years ago2
Training hyperparameters？
Closed 3 years ago1
About TAggreagate
Closed 3 years ago1
from .layers.drop import DropPath
Closed 3 years ago2