search

landskape-ai / triplet-attention

Official PyTorch Implementation for "Rotate to Attend: Convolutional Triplet Attention Module." [WACV 2021]
https://openaccess.thecvf.com/content/WACV2021/html/Misra_Rotate_to_Attend_Convolutional_Triplet_Attention_Module_WACV_2021_paper.html
MIT License
406 stars 49 forks source link

预训练权重 #2

Closed ZJUTSong closed 4 years ago

ZJUTSong commented 4 years ago

我这边使用res2net的话，看到您这边没有对应的res2net+triplet_attentation预训练权重，会有影响吗？

digantamisra98 commented 4 years ago

Hi, we didn't train a Res2Net with triplet attention and therefore don't have any corresponding training weights. We plan to train Efficient Net and Res2Net backbones ahead in the near future. However, we are strongly confident that training a Res2Net with Triplet attention won't cause any issues and should provide performance boost as highlighted by our existing results. Feel free to notify us in case of any issues regarding Triplet's performance in Res2Net. Thanks!