赞
踩
目录
1.Demystify Transformers & Convolutions in Modern Image Deep Networks
1.2 自注意力模型:Halo Attention
2 .HaloAttn引入到Yolov5/Yolov7