sgrvinod / a-PyTorch-Tutorial-to-Object-Detection

SSD: Single Shot MultiBox Detector | a PyTorch Tutorial to Object Detection
MIT License
3.05k stars 719 forks source link

Wrong feature map size after conv7 in the picture #100

Open denisshaf opened 9 months ago

denisshaf commented 9 months ago

The size of the feature map after the 4th max-pooling is 19x19, then it's (19 - 3) / 1 + 1 = 17 after 3x3 pooling with stride 1, and after conv6 the size is (17 + 26 - (62 + 1)) / 1 + 1 = 17, but the size in the picture is 19.