Faster rcnn--改变anchor的size

Faster rcnn–改变anchor的size

• 代码用的是py-faster-rcnn：https://github.com/rbgirshick/py-faster-rcnn
• 训练时用的模型和方法：models\pascal_voc\VGG_CNN_M_1024\faster_rcnn_end2end

---

修改generate_anchors.py

#ratios中加了一种比例1.5，所以anchor的数量为4*3=12
def generate_anchors(base_size=16, ratios=[0.5, 1, 1.5, 2],
                     scales=2**np.arange(3, 6)):
    """
    Generate anchor (reference) windows by enumerating aspect ratios X
    scales wrt a reference (0, 0, 15, 15) window.
    """

    base_anchor = np.array([1, 1, base_size, base_size]) - 1
    ratio_anchors = _ratio_enum(base_anchor, ratios)
    anchors = np.vstack([_scale_enum(ratio_anchors[i, :], scales)
                         for i in xrange(ratio_anchors.shape[0])])
    return anchors

修改train.prototxt

layer {
  name: "rpn_cls_score"
  type: "Convolution"
  bottom: "rpn/output"
  top: "rpn_cls_score"
  param { lr_mult: 1.0 }
  param { lr_mult: 2.0 }
  convolution_param {
    num_output: 24   # 2(bg/fg) * 12(anchors)
    kernel_size: 1 pad: 0 stride: 1
    weight_filler { type: "gaussian" std: 0.01 }
    bias_filler { type: "constant" value: 0 }
  }

layer {
  name: "rpn_bbox_pred"
  type: "Convolution"
  bottom: "rpn/output"
  top: "rpn_bbox_pred"
  param { lr_mult: 1.0 }
  param { lr_mult: 2.0 }
  convolution_param {
    num_output: 48   # 4 * 12(anchors)
    kernel_size: 1 pad: 0 stride: 1
    weight_filler { type: "gaussian" std: 0.01 }
    bias_filler { type: "constant" value: 0 }
  }
}

layer {
  name: 'rpn_cls_prob_reshape'
  type: 'Reshape'
  bottom: 'rpn_cls_prob'
  top: 'rpn_cls_prob_reshape'
  reshape_param { shape { dim: 0 dim: 24 dim: -1 dim: 0 } }
}

修改test.prototxt

同train.prototxt

上述步骤只针对改ratios，如果你想改scales，还要进行以下修改

修改generate_anchors.py

#anchor的数量为4*5=20
def generate_anchors(base_size=16, ratios=[0.5, 1, 1.5, 2],
                     scales=2**np.arange(1, 6)):
    """
    Generate anchor (reference) windows by enumerating aspect ratios X
    scales wrt a reference (0, 0, 15, 15) window.
    """

    base_anchor = np.array([1, 1, base_size, base_size]) - 1
    ratio_anchors = _ratio_enum(base_anchor, ratios)
    anchors = np.vstack([_scale_enum(ratio_anchors[i, :], scales)
                         for i in xrange(ratio_anchors.shape[0])])
    return anchors

修改anchor_target_layer.py

    def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str)
        anchor_scales = layer_params.get('scales', (2, 4, 8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

修改proposal_layer.py

    def setup(self, bottom, top):
        # parse the layer parameter string, which must be valid YAML
        layer_params = yaml.load(self.param_str)

        self._feat_stride = layer_params['feat_stride']
        anchor_scales = layer_params.get('scales', (2, 4, 8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]

如果报错，删除faster_rcnn_version\fasterRcnn_final_v1_3\data\cache下.pkl文件

---