MATLAB Answers

Imported Net from Caffe Missing Classification Labels

1 view (last 30 days)
Miraç  Gülgönül
Miraç Gülgönül on 12 Sep 2019
Answered: Shivam Sardana on 24 Sep 2019
When I import a pre-trained Caffee net with importCaffeNetwork including the .prototxt and .caffemodel files and look at the last layer to see the classification labels it outputs:
ClassificationOutputLayer with properties:
Name: 'ClassificationOutput'
Classes: [class1 class2 class3 class4 class5 class6 class7]
OutputSize: 7
Hyperparameters
LossFunction: 'crossentropyex'
I would assume that these labels are included in the files, so why are they missing here?

  0 Comments

Sign in to comment.

Answers (1)

Shivam Sardana
Shivam Sardana on 24 Sep 2019
There is a name-apir input 'Classes' argument which you can pass in importCaffeNetwork. You can refer to this documentation link for more detail:
Hope this help.

  0 Comments

Sign in to comment.

Sign in to answer this question.

Products


Release

R2019a