Skip to main content
Data Science

Return to Answer

added 2 characters in body
Source Link
Ethan
Ethan
  • 1.7k
  • 9
  • 24
  • 39

The error is simple when you input a blank image , lets. Let us say black image just like you have done with tf.zerostf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv)sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be 0 and thus will not change  . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same  . Thus it will not learn anything  .

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be 0 and thus will not change  . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same  . Thus it will not learn anything  .

The error is simple when you input a blank image. Let us say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be 0 and thus will not change. No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same. Thus it will not learn anything.

added 2 characters in body
Source Link
Lucid
Lucid
  • 111
  • 3

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be 0 and thus will not change . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same . Thus it will not learn anything .

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be and thus will not change . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same . Thus it will not learn anything .

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be 0 and thus will not change . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same . Thus it will not learn anything .

deleted 26 characters in body
Source Link
Lucid
Lucid
  • 111
  • 3

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is 0 multiplying it by any conv filled with 0 the dot product will be 0 sum it up 0 .

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be and thus will not change . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same . Thus it will not learn anything .

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is 0 multiplying it by any conv filled with 0 will be 0 sum it up 0 .

Then we will add the bias every value in the resulting image will be the bias activation next still the same value . No matter what input you will give and what weights you will have in the first layer its gradient will be and thus will not change . No matter what input image you give the resulting image will be filled with the bias values and the final output will be the same . Thus it will not learn anything .

The error is simple when you input a blank image , lets say black image just like you have done with tf.zeros the resulting images will be full of 0s so coming to the point when you will multiply your weights by 0 the resulting dot product will be 0

because the formula is sum(weights * conv) and as conv is filled with 0 the dot product will be 0.

Then we will add the bias every value in the resulting image will be the bias activation next still the same value everywhere. No matter what input you will give and what weights you will have in the first layer its gradient will be and thus will not change . No matter what input image you give the resulting image will be 0 filled with the bias values and the final output will be the same . Thus it will not learn anything .

Source Link
Lucid
Lucid
  • 111
  • 3