Problem with RCNN traning and testing..

[zuacubd]

I got the following errors when I try to run rcnn training and testing on PASCAL VOC 2012 dataset.
test_results = rcnn_exp_train_test();

Error using *
Inputs must be 2-D, or at least one input must be scalar.
To compute elementwise TIMES, use TIMES (.*) instead.

Error in rcnn_pool5_to_fcX (line 32)
feat = max(0, bsxfun(@plus, feat*rcnn_model.cnn.layers(i).weights{1}, ...

Error in rcnn_train (line 101)
X_pos{i} = rcnn_pool5_to_fcX(X_pos{i}, opts.layer, rcnn_model);

Error in rcnn_exp_train_and_test (line 24)
[rcnn_model, rcnn_k_fold_model] = ...

It seems to me that dimensions of feat and weights{1} are not compatible.
It my program, the dimensions of

size(feat) = 954 9216
size(rcnn_model.cnn.layers(i).weights{1}) = 1 1 9216 4096
size(rcnn_model.cnn.layers(i).weights{2}) = 1 1 1 4096

[zuacubd]

Problem solved. It was Caffe version.
RCNN is compatible with Caffe-0.999 version.