Data shape in convolutional layer with data sets such as intestinal parasites and plasmodium (motic)

[vinayakumarr]

intestinal parasites

In convolutional Layer, how you are getting

7x56x56 (it should be 7x58x58) - intestinal parasites

7x21x21 (it should be 7x23x23) - plasmodium (motic)

intestinal parasites

Layer information

name size

---

0 input 3x60x60 1 conv1 7x56x56 ((60-3)/1) + 1 = 58 (how 56?) 2 pool1 7x28x28 3 conv2 12x27x27 4 hidden3 500 5 output 2

tuberculosis

Layer information

name size

---

0 input 3x20x20 1 conv1 7x18x18 ((20-3)/1) + 1 = 18 2 pool1 7x9x9 3 conv2 12x8x8 4 hidden3 100 5 output 2

plasmodium (motic)

Layer information

name size

---

0 input 3x25x25 1 conv1 7x21x21 ((25-3)/1) + 1 = 23 (how 21?) 2 pool1 7x10x10 3 conv2 12x9x9 4 hidden3 500 5 output 2

plasmodium (phone)

Layer information

name size

---

0 input 3x20x20 1 conv1 7x18x18 ((20-3)/1) + 1 = 18 2 pool1 7x9x9 3 conv2 12x8x8 4 hidden3 500 5 output 2