Faster RCNN训练出现问题：Matlab版本训练

I1113 20:56:49.405777 23834 net.cpp:42] Initializing net from parameters: 
name: "Zeiler_conv5"
input: "data"
input_dim: 1
input_dim: 3
input_dim: 224
input_dim: 224
state {
  phase: TEST
}
layer {
  name: "conv1"
  type: "Convolution"
  bottom: "data"
  top: "conv1"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 96
    pad: 3
    kernel_size: 7
    stride: 2
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "relu1"
  type: "ReLU"
  bottom: "conv1"
  top: "conv1"
}
layer {
  name: "norm1"
  type: "LRN"
  bottom: "conv1"
  top: "norm1"
  lrn_param {
    local_size: 3
    alpha: 5e-05
    beta: 0.75
    norm_region: WITHIN_CHANNEL
  }
}
layer {
  name: "pool1"
  type: "Pooling"
  bottom: "norm1"
  top: "pool1"
  pooling_param {
    pool: MAX
    kernel_size: 3
    stride: 2
    pad: 1
  }
}
layer {
  name: "conv2"
  type: "Convolution"
  bottom: "pool1"
  top: "conv2"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 256
    pad: 2
    kernel_size: 5
    stride: 2
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 1
    }
  }
}
layer {
  name: "relu2"
  type: "ReLU"
  bottom: "conv2"
  top: "conv2"
}
layer {
  name: "norm2"
  type: "LRN"
  bottom: "conv2"
  top: "norm2"
  lrn_param {
    local_size: 3
    alpha: 5e-05
    beta: 0.75
    norm_region: WITHIN_CHANNEL
  }
}
layer {
  name: "pool2"
  type: "Pooling"
  bottom: "norm2"
  top: "pool2"
  pooling_param {
    pool: MAX
    kernel_size: 3
    stride: 2
    pad: 1
  }
}
layer {
  name: "conv3"
  type: "Convolution"
  bottom: "pool2"
  top: "conv3"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 384
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "relu3"
  type: "ReLU"
  bottom: "conv3"
  top: "conv3"
}
......
I1113 20:56:49.433302 23834 net.cpp:194] relu5 does not need backward computation.
I1113 20:56:49.433305 23834 net.cpp:194] conv5 does not need backward computation.
I1113 20:56:49.433307 23834 net.cpp:194] relu4 does not need backward computation.
I1113 20:56:49.433310 23834 net.cpp:194] conv4 does not need backward computation.
I1113 20:56:49.433312 23834 net.cpp:194] relu3 does not need backward computation.
I1113 20:56:49.433315 23834 net.cpp:194] conv3 does not need backward computation.
I1113 20:56:49.433317 23834 net.cpp:194] pool2 does not need backward computation.
I1113 20:56:49.433320 23834 net.cpp:194] norm2 does not need backward computation.
I1113 20:56:49.433323 23834 net.cpp:194] relu2 does not need backward computation.
I1113 20:56:49.433326 23834 net.cpp:194] conv2 does not need backward computation.
I1113 20:56:49.433328 23834 net.cpp:194] pool1 does not need backward computation.
I1113 20:56:49.433331 23834 net.cpp:194] norm1 does not need backward computation.
I1113 20:56:49.433334 23834 net.cpp:194] relu1 does not need backward computation.
I1113 20:56:49.433336 23834 net.cpp:194] conv1 does not need backward computation.
I1113 20:56:49.433339 23834 net.cpp:235] This network produces output proposal_bbox_pred
I1113 20:56:49.433341 23834 net.cpp:235] This network produces output proposal_cls_prob
I1113 20:56:49.433354 23834 net.cpp:492] Collecting Learning Rate and Weight Decay.
I1113 20:56:49.433359 23834 net.cpp:247] Network initialization done.
I1113 20:56:49.433362 23834 net.cpp:248] Memory required for data: 21358056

错误：
Preparing training data...Starting parallel pool (parpool) using the 'local' profile ... connected to 4 workers.
Done.
Preparing validation data...Done.
Error using randperm
K must be less than or equal to N.


Error in proposal_train (line 88)
        shuffled_inds_val   = shuffled_inds_val(randperm(length(shuffled_inds_val), opts.val_iters));


Error in Faster_RCNN_Train.do_proposal_train (line 7)
    model_stage.output_model_file = proposal_train(conf, dataset.imdb_train, dataset.roidb_train, ...


Error in script_faster_rcnn_VOC2007_ZF (line 45)
model.stage1_rpn            = Faster_RCNN_Train.do_proposal_train(conf_proposal, dataset, model.stage1_rpn, opts.do_val);
解决：
functions/fast_rcnn/fast_rcnn_train.m 
将val_iters改小，为val的1/5

转载请注明：http://blog.csdn.net/forest_world