Py-faster-rcnn: How to train a ZF network on one class (person) getting ValueError: could not broadcast input array from shape (4) into shape (0)

[emuchogu]

I'm trying to train a ZF network on one class "person" but I'm getting the following error when I run the code:

File "/home/edward/caffee/r-cnn/faster-rcnn/py-faster-rcnn/tools/../lib/rpn/proposal_target_layer.py", line 138, in _get_bbox_regression_labels
bbox_targets[ind, start:end] = bbox_target_data[ind, 1:]
ValueError: could not broadcast input array from shape (4) into shape (0)

I initialize the training process by running the following command:
./experiments/scripts/faster_rcnn_end2end.sh 0 ZF

I changed the following files:
models/ZF/faster_rcnn_end2end/train.prototxt

---

name: "ZF"
layer {
name: 'input-data'
type: 'Python'
top: 'data'
top: 'im_info'
top: 'gt_boxes'
python_param {
module: 'roi_data_layer.layer'
layer: 'RoIDataLayer'
param_str: "'num_classes': 21"
}
}

to

name: "ZF"
layer {
name: 'input-data'
type: 'Python'
top: 'data'
top: 'im_info'
top: 'gt_boxes'
python_param {
module: 'roi_data_layer.layer'
layer: 'RoIDataLayer'
param_str: "'num_classes': 2"
}

}

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: 21
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}

to

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: 2
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: 84
weight_filler {
type: "gaussian"
std: 0.001
}
bias_filler {
type: "constant"
value: 0
}
}
}

to

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: 8
weight_filler {
type: "gaussian"
std: 0.001
}
bias_filler {
type: "constant"
value: 0
}
}

}

I also changed :

lib/datasets/pascal_voc.py
Line 28:

---

    self._classes = ('__background__', # always index 0
                     'person', 'bicycle', 'bird', 'boat',
                     'bottle', 'bus', 'car', 'cat', 'chair',
                     'cow', 'diningtable', 'dog', 'horse',
                     'motorbike', 'person', 'pottedplant',
                     'sheep', 'sofa', 'train', 'tvmonitor')

to

    self._classes = ('__background__', # always index 0
                     'person')

---

I would like to document how to train a faster r-cnn network on a single class, by documenting how a single class is trained I and other should then be able to more easily adapted the code to train multiple classes as well as train using our own data-sets.

[wait1988]

Have you solved the problem?I just got the same error.

[emuchogu]

Nope. Can you share details on how you have implemented yours, maybe that way we can work on it together and find a solution.

[longchuanshu]

Have you solved the problem?I just got the same error.

[shoCaffeTeria]

I have the same error,too. Could you cope with that?
I tryed it with VGG_CNN_M_1024 but it couldn't work. The same error happened.

[shoCaffeTeria]

I think I could solve the error.
In pascal_voc.py(202)
from
cls = self._class_to_ind[obj.find('name').text.lower().strip()]
to
cls = self._class_to_ind['person']

[emuchogu]

Did changing from
cls = self._class_to_ind[obj.find('name').text.lower().strip()]
to
cls = self._class_to_ind['person']

Work?

On Thu, Mar 17, 2016 at 6:10 AM, shoCaffeTeria [email protected]
wrote:

I think I could solve the error.
In pascal_voc.py(202)
from
cls = self._class_to_ind[obj.find('name').text.lower().strip()]
to
cls = self._class_to_ind['person']

—
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub
https://github.com/rbgirshick/py-faster-rcnn/issues/52#issuecomment-197672199

[shoCaffeTeria]

Yes, It worked well.

[ednarb29]

How can that be handled with more classes and what's with the background class?

[JohnnyY8]

I have the same error as well.
But changing from cls = self._class_to_ind[obj.find('name').text.lower().strip()]
to cls = self._class_to_ind['person'] is not good for me.
Because I want to train faster r-cnn on DET dataset, can anyone help me ?

[macqueen09]

same question after I use my own dataset which has 3 class . confused me is that I use this dataset to train a faster RCNN network success last days , after some experience , I come back this train , and It break , I'm confused

[tony5614]

you can try to remove ./data/cache/voc_2007_trainval_gt_roidb.pkl

find -name "*roidb.pkl" |xargs rm

It works for me

[kecaiwu]

how to solve this error #486 ? @shoCaffeTeria

[clvoloshin]

Check that solver.prototxt points to the correct train/test prototxt