kanezaki/rotationnet

About the initial weights

Opened this issue · 8 comments

lh641446825 commented

Sorry for bothering you. I download initial weights for fine-tuning the models, but when I unzip the package, I got the following information:

gzip: stdin: not in gzip format
tar: Child returned status 1
tar: Error is not recoverable: exiting now

I found that the website http://www.cs.berkeley.edu/~rbg/r-cnn-release1-data.tgz did not have the r-cnn-release1-data.tgz package. Can you provide the updated website?

lh641446825 commented

Thanks! I downloaded the weights. When I input the command /home/caffe-rotationnet2/build/tools/caffe train -solver Training/rotationnet_modelnet40_case2_solver.prototxt -weights caffe_nets/ilsvrc_2012_train_iter_310k 2>&1 | tee log.txt, I got the following information:

F0418 08:04:43.886162 6901 layer_factory.hpp:81] Check failed: registry.count(type) == 1 (0 vs. 1) Unknown layer type: ImageData (known types: AbsVal, Accuracy, ArgMax, BNLL, BatchNorm, BatchReindex, Bias, Concat, ContrastiveLoss, Convolution, Crop, Data, Deconvolution, Dropout, DummyData, ELU, Eltwise, Embed, EuclideanLoss, Exp, Filter, Flatten, HDF5Data, HDF5Output, HingeLoss, Im2col, InfogainLoss, InnerProduct, Input, LRN, LSTM, LSTMUnit, Log, MVN, MemoryData, MultinomialLogisticLoss, MyAccuracy, MySoftmax, MySoftmaxLoss, MySoftmaxWithLoss, PReLU, Parameter, Pooling, Power, Python, RNN, ReLU, Reduction, Reshape, SPP, Scale, Sigmoid, SigmoidCrossEntropyLoss, Silence, Slice, Softmax, SoftmaxWithLoss, Split, TanH, Threshold, Tile)
*** Check failure stack trace: ***
@ 0x7f675a802daa (unknown)
@ 0x7f675a802ce4 (unknown)
@ 0x7f675a8026e6 (unknown)
@ 0x7f675a805687 (unknown)
@ 0x7f675ae33774 caffe::LayerRegistry<>::CreateLayer()
@ 0x7f675ae38d0c caffe::Net<>::Init()
@ 0x7f675ae3ae05 caffe::Net<>::Net()
@ 0x7f675af6ceaa caffe::Solver<>::InitTrainNet()
@ 0x7f675af6dfac caffe::Solver<>::Init()
@ 0x7f675af6e2da caffe::Solver<>::Solver()
@ 0x7f675af52f63 caffe::Creator_SGDSolver<>()
@ 0x40f48e caffe::SolverRegistry<>::CreateSolver()
@ 0x4084e4 train()
@ 0x405e4c main
@ 0x7f6759573f45 (unknown)
@ 0x406753 (unknown)
@ (nil) (unknown)

What I should do?

kanezaki commented

I'm not sure about that error... Have you compiled caffe-rotationnet2 with no error?

lh641446825 commented

Yes, I compiled caffe-rotationnet2 with no error. I got the following information:

CXX/LD -o .build_release/examples/cpp_classification/classification.bin
CXX examples/cifar10/convert_cifar_data.cpp
CXX/LD -o .build_release/examples/cifar10/convert_cifar_data.bin
CXX examples/siamese/convert_mnist_siamese_data.cpp
CXX/LD -o .build_release/examples/siamese/convert_mnist_siamese_data.bin
CXX/LD -o python/caffe/_caffe.so python/caffe/_caffe.cpp
touch python/caffe/proto/init.py
PROTOC (python) src/caffe/proto/caffe.proto

kanezaki commented

Understood. These lines in Makefile.config should be left commented out.

# uncomment to disable IO dependencies and corresponding data layers
# USE_OPENCV := 0
# USE_LEVELDB := 0
# USE_LMDB := 0
lh641446825 commented

Sorry. I comment out # USE_OPENCV := 0, then compile caffe-rotationnet2. I got this error :

LD -o .build_release/lib/libcaffe.so.1.0.0-rc3
/usr/bin/ld: cannot find -lopencv_imgcodecs
collect2: error: ld returned 1 exit status
make: *** [.build_release/lib/libcaffe.so.1.0.0-rc3] Error 1

I can't find how to solve it. In this issue someone said uncomment USE_OPENCV := 0

lh641446825 commented

Sorry, I have installed OpenCV before.