Geek Answers Handbook

Caffe net.predict () prints random results (GoogleNet)

I used GoogleNet pre-validation from https://github.com/BVLC/caffe/tree/master/models/bvlc_googlenet and finalized it with my own data (~ 100K images, 101 class). After one day of training, I reached 62% in the top 1 and 85% in the top 5 classification and tried to use this network to predict multiple images.

I just followed the example from https://github.com/BVLC/caffe/blob/master/examples/classification.ipynb ,

Here is my Python code:

import caffe import numpy as np caffe_root = './caffe' MODEL_FILE = 'caffe/models/bvlc_googlenet/deploy.prototxt' PRETRAINED = 'caffe/models/bvlc_googlenet/bvlc_googlenet_iter_200000.caffemodel' caffe.set_mode_gpu() net = caffe.Classifier(MODEL_FILE, PRETRAINED, mean=np.load('ilsvrc_2012_mean.npy').mean(1).mean(1), channel_swap=(2,1,0), raw_scale=255, image_dims=(224, 224)) def caffe_predict(path): input_image = caffe.io.load_image(path) print path print input_image prediction = net.predict([input_image]) print prediction print "----------" print 'prediction shape:', prediction[0].shape print 'predicted class:', prediction[0].argmax() proba = prediction[0][prediction[0].argmax()] ind = prediction[0].argsort()[-5:][::-1] # top-5 predictions return prediction[0].argmax(), proba, ind

In my deploy.prototxt file, I changed the last layer just to predict my 101 classes.

 layer { name: "loss3/classifier" type: "InnerProduct" bottom: "pool5/7x7_s1" top: "loss3/classifier" param { lr_mult: 1 decay_mult: 1 } param { lr_mult: 2 decay_mult: 0 } inner_product_param { num_output: 101 weight_filler { type: "xavier" } bias_filler { type: "constant" value: 0 } } } layer { name: "prob" type: "Softmax" bottom: "loss3/classifier" top: "prob" }

Here is the distribution of softmax output:

 [[ 0.01106235 0.00343131 0.00807581 0.01530041 0.01077161 0.0081002 0.00989228 0.00972753 0.00429183 0.01377776 0.02028225 0.01209726 0.01318955 0.00669979 0.00720005 0.00838189 0.00335461 0.01461464 0.01485041 0.00543212 0.00400191 0.0084842 0.02134697 0.02500303 0.00561895 0.00776423 0.02176422 0.00752334 0.0116104 0.01328687 0.00517187 0.02234021 0.00727272 0.02380056 0.01210031 0.00582192 0.00729601 0.00832637 0.00819836 0.00520551 0.00625274 0.00426603 0.01210176 0.00571806 0.00646495 0.01589645 0.00642173 0.00805364 0.00364388 0.01553882 0.01549598 0.01824486 0.00483241 0.01231962 0.00545738 0.0101487 0.0040346 0.01066607 0.01328133 0.01027429 0.01581303 0.01199994 0.00371804 0.01241552 0.00831448 0.00789811 0.00456275 0.00504562 0.00424598 0.01309276 0.0079432 0.0140427 0.00487625 0.02614347 0.00603372 0.00892296 0.00924052 0.00712763 0.01101298 0.00716757 0.01019373 0.01234141 0.00905332 0.0040798 0.00846442 0.00924353 0.00709366 0.01535406 0.00653238 0.01083806 0.01168014 0.02076091 0.00542234 0.01246306 0.00704035 0.00529556 0.00751443 0.00797437 0.00408798 0.00891858 0.00444583]]

This is like random distribution without meaning.

Thanks for any help or hint and best wishes, Alex

+5
source share
2 answers

The solution is really simple: I just forgot to rename the last layer in the deployment file:

 layer { name: "loss3/classifier" type: "InnerProduct" bottom: "pool5/7x7_s1" top: "loss3/classifier" param { lr_mult: 1 decay_mult: 1 }
+2
source

Please check the image conversion you are using - is it the same in training and testing?

AFAIK bvlc_googlenet subtracts the average value of an image with one value per channel, while your python classifier uses a different average mean=np.load('ilsvrc_2012_mean.npy').mean(1).mean(1) . This can cause the network to not be able to correctly classify your inputs.

0
source

More articles:


All Articles