search

ini17 / Micro-expression-recognition-by-fusing-Action-Unit-detection-and-spatio-temporal-features

Codes for our paper -- Micro-expression recognition by fusing Action Unit detection and spatio-temporal features.
2 stars 1 forks source link

MagNet的预训练参数 #2

Open ANGLE404 opened 2 weeks ago

ANGLE404 commented 2 weeks ago 
RIFE = Model()  # 创建RIFE模型实例
RIFE.load_model(r"B:\0_0NewLife\0_Papers\FGRMER\weight", -1)  # 加载模型权重
RIFE.eval()  # 设置模型为评估模式
RIFE.device()  # 设置模型使用的设备

您在readme内删除了

MagNet

The structure of MagNet was adapted from here. Please download the pretrained weight from their release and place in dataloader/weight/.

这一部分，是不是不需要下载参数了，我在您给的地址内找到了参数 magnet_epoch12_loss7.28e-02.pth ，请问是这个参数吗？

https://github.com/ZhengPeng7/motion_magnification_learning-based/releases/tag/v1.0

ini17 commented 2 weeks ago

是的，就是这个预训练模型😊😊

ANGLE404 commented 1 week ago

感谢您的回复