ASF-YOLO
ASF-YOLO copied to clipboard
有一个论文问题想请教一下作者
"In SSFF, The P3, P4, and P5 feature maps are normalized to the same size, upsampled, and then stacked together as input to a 3D convolution to combine multiscale features." See Page 6.
Thanks for pointing this problem out. To clarify the expression, we replace scale-invariant features' with
aspect ratio invariant features'. As you know, each learned filter of Convolutional neural networks (CNNs), including upsampling and downsampling, is sensitive to a given set of features only within a narrow range of scale. During the process, all images are resized, but the aspect ratio, which is the ratio of their width to height, is preserved.