折腾faster-rcnn（四）--测试篇

标签（空格分隔）： Linux 深度学习

1.修改测试脚本：py-faster-rcnn/models/pascal_voc/ZF/faster_rcnn_alt_opt/test.pt

layer {
  name: "cls_score"
  type: "InnerProduct"
  bottom: "fc7"
  top: "cls_score"
  param { lr_mult: 1.0 }
  param { lr_mult: 2.0 }
  inner_product_param {
    num_output: 5 #按训练集类别改，该值为类别数+1
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}

layer {
  name: "bbox_pred"
  type: "InnerProduct"
  bottom: "fc7"
  top: "bbox_pred"
  param { lr_mult: 1.0 }
  param { lr_mult: 2.0 }
  inner_product_param {
    num_output: 20 #按训练集类别改，该值为（类别数+1）*4
    weight_filler {
      type: "gaussian"
      std: 0.001
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}

2.执行测试代码：

./tools/test_net.py --gpu 0 --def models/pascal_voc/ZF/faster_rcnn_alt_opt/faster_rcnn_test.pt --net data/faster_rcnn_models/ZF_faster_rcnn_final.caffemodel --imdb voc_2007_test --cfg experiments/cfgs/faster_rcnn_alt_opt.yml

3.发现报错

IOError: [Errno 2] No such file or directory: '/home/wangzhan/py-faster-rcnn-master/data/VOCdevkit2007/results/VOC2007/Main/comp4_e3ae962b-98ad-418e-a396-bc6fa4d1d62f_det_test_kiss.txt'

这是由于VOCdevkit2007目录下只有VOC2007目录，没有results目录。构建数据集的工作请参阅折腾faster-rcnn（三）--训练篇的相关部分。
4.再次运行步骤2中代码，得到测试集中每一类的AP（Average Precision）和这4类的mAP：

AP for kiss = 0.7187
AP for hug = 0.5884
AP for hands = 0.4475
AP for person = 0.7523
Mean AP = 0.6267
~~~~~~~~
Results:
0.719
0.588
0.447
0.752
0.627
~~~~~~~~