caffe --- mistakes

1. Check failed: img_height == data_mean_.height() (224 vs. 256)

layer {

  name: "data"
  type: "ImageData"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
  transform_param {
    mirror: true
    crop_size: 224
    mean_file: "mean_sub.binaryproto"    # 256
  }
  image_data_param {
    source: "triplet_sampled_train.txt"      #  加路径名
    batch_size: 100
    new_height: 256      #  It will also resize images
    new_width: 256
    shuffle: true         # 每个epoch shuffle
    pair_size: 2
  }

}

2. mean

name.mean(1).mean(1)  #是一个三个元素的数组, 每个通道一个值...    name shape eg. 3X256X256