Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.
6404 Discussions

Batch normalization: [Error 5] Toolkit Error: Stage Details Not Supported: Neg

idata
Employee
706 Views

Hello,

 

I am trying to use the batch normalization layer from tensorflow, but the mvNCCompile (NCSDK version 2.08.01) would not complete because it says: [Error 5] Toolkit Error: Stage Details Not Supported: Neg.

 

I did look into why this happens and the batch normalize layer is actually split into the following nodes:

 

node.name= batchnorm/add/y node.type= Const

 

node.name= batchnorm/add node.type= Add

 

node.name= batchnorm/Rsqrt node.type= Rsqrt

 

node.name= batchnorm/mul node.type= Mul

 

node.name= batchnorm/mul_1 node.type= Mul

 

node.name= batchnorm/Neg node.type= Neg

 

Where the last one is indeed a Neg operation which, as far as I know, is simply a sign inversion of each value.

 

My concern is that it is said on the release page that the batch normalization layer should be supported.

 

Here is how I use the batch normalization layer, where conv is the previous layer, and means, variances and scales are the weights of the batch normalization layer, of type tf.Variable:

 

conv = tf.nn.batch_normalization( x=conv, mean=means, variance=variances, offset=None, scale=scales, variance_epsilon=1e-5 )

 

Any clue?

 

Thank you very much!
0 Kudos
0 Replies
Reply