Incompatible with Keras > 2.0 #3

SaschaHornauer · 2017-03-22T06:51:48Z

The new API of any Keras version bigger than 2.0 seems to lead to unconnected networks. No error of any kind can be seen apart from that the model.summary() will not show any "Connected to" entry and the network will not find any vehicle.

Downgrading to version 1.2.2. resolves the issue.

Thanks for the great work creating this, I would fix the issue myself though I am not yet familiar enough with Keras and do not yet know how.

xslittlegrass · 2017-03-27T17:57:00Z

Thanks for letting me know. I will fix this when I figure out what's are changes of the API in the new version.

im2ex · 2017-04-27T13:28:18Z

Why is this marked closed?
I ran into same problem.
Since there is no error i spent quite some time trying to find out whats going on.
Everything seems to be ok, but no boxes. With small threshold there are boxes but only wrong ones.
I did look into the open Issues early, but it took me long to find @SaschaHornauer solution (Thanks a lot!).
Besides opening the issue again, I would suggest to add a heads up comment in the script (eg. import keras % broken for keras >= 2.0, use 1.2.2)

Otherwise: great ipynb! Learned a lot from this. Very compact and clear. Thanks.

xslittlegrass · 2017-04-27T15:01:14Z

@im2ex I'll add the comment, thanks for the reminder.

liborw · 2017-05-05T08:57:54Z

I had same problem, so I downgraded to keras=1.2.2 then saved entire model with model.save_model, then upgrade to keras=2.0.4 load the model and it works fine...with some warning to migrate to new API.

marc-chan · 2017-07-27T04:16:56Z

This issue is due to the difference in the conv layer weights implementation in Keras 1 and 2.

In Keras 2, weights are stored in the same way regardless of the image dimension ordering set ('th' or 'tf') - channels last: i.e. (3,3,3,16) for a conv layer with a 3x3 kernel, 3 input channels and 16 output channels.

Whereas in Keras 1, depending on the image dimension ordering set, the conv layer weights take on different dimensions, wrt to above example: (3,3,3,16) for 'tf' and (16,3,3,3) for 'th'.

A bit ugly, but code below should work for Keras 2:

p.s. Remember to set keras.backend.set_image_dim_ordering('th') for Keras 2 as well....

def load_weights(model,yolo_weight_file):           
    
    tiny_data = np.fromfile(yolo_weight_file,np.float32)[4:]
  
    index = 0
    for layer in model.layers:
        weights = layer.get_weights()
        if len(weights)>0:
            filter_shape, bias_shape = [w.shape for w in weights]
            if len(filter_shape)>2: #For convolutional layers
                filter_shape_i = filter_shape[::-1]
                bias_weight = tiny_data[index:index+np.prod(bias_shape)].reshape(bias_shape)
                index += np.prod(bias_shape)
                filter_weight= tiny_data[index:index+np.prod(filter_shape_i)].reshape(filter_shape_i)
                filter_weight= np.transpose(filter_weight,(2,3,1,0))
                index += np.prod(filter_shape)
                layer.set_weights([filter_weight,bias_weight])
            else: #For regular hidden layers
                bias_weight = tiny_data[index:index+np.prod(bias_shape)].reshape(bias_shape)
                index += np.prod(bias_shape)
                filter_weight= tiny_data[index:index+np.prod(filter_shape)].reshape(filter_shape)
                index += np.prod(filter_shape)
                layer.set_weights([filter_weight,bias_weight])

sungmin-yang · 2017-12-06T12:09:08Z

@marc-chan Thanks for your help!

XiongLianga · 2018-03-29T14:42:25Z

Thanks for your solution

Alteregoxxx · 2018-10-09T13:22:21Z

This issue is due to the difference in the conv layer weights implementation in Keras 1 and 2.

In Keras 2, weights are stored in the same way regardless of the image dimension ordering set ('th' or 'tf') - channels last: i.e. (3,3,3,16) for a conv layer with a 3x3 kernel, 3 input channels and 16 output channels.

Whereas in Keras 1, depending on the image dimension ordering set, the conv layer weights take on different dimensions, wrt to above example: (3,3,3,16) for 'tf' and (16,3,3,3) for 'th'.

A bit ugly, but code below should work for Keras 2:

p.s. Remember to set keras.backend.set_image_dim_ordering('th') for Keras 2 as well....
def load_weights(model,yolo_weight_file):           
    
    tiny_data = np.fromfile(yolo_weight_file,np.float32)[4:]
  
    index = 0
    for layer in model.layers:
        weights = layer.get_weights()
        if len(weights)>0:
            filter_shape, bias_shape = [w.shape for w in weights]
            if len(filter_shape)>2: #For convolutional layers
                filter_shape_i = filter_shape[::-1]
                bias_weight = tiny_data[index:index+np.prod(bias_shape)].reshape(bias_shape)
                index += np.prod(bias_shape)
                filter_weight= tiny_data[index:index+np.prod(filter_shape_i)].reshape(filter_shape_i)
                filter_weight= np.transpose(filter_weight,(2,3,1,0))
                index += np.prod(filter_shape)
                layer.set_weights([filter_weight,bias_weight])
            else: #For regular hidden layers
                bias_weight = tiny_data[index:index+np.prod(bias_shape)].reshape(bias_shape)
                index += np.prod(bias_shape)
                filter_weight= tiny_data[index:index+np.prod(filter_shape)].reshape(filter_shape)
                index += np.prod(filter_shape)
                layer.set_weights([filter_weight,bias_weight])

@marc-chan
Hello, sorry to resurrect this old post, but your help would be very very appreciated :-)
I've tried to use your alternative function for loading the weights, still, the problems remain the same: No bounding boxes, unless one lower the threshold almost to zero...(they are in the wrong position, nevertheless)
Oh, obviously I did not forget to use

keras.backend.set_image_dim_ordering('th')

I'm using Keras 2.2.4 version.

Thank you!

Wentaobi · 2018-10-11T04:12:53Z

This issue is due to the difference in the conv layer weights implementation in Keras 1 and 2.

In Keras 2, weights are stored in the same way regardless of the image dimension ordering set ('th' or 'tf') - channels last: i.e. (3,3,3,16) for a conv layer with a 3x3 kernel, 3 input channels and 16 output channels.

Whereas in Keras 1, depending on the image dimension ordering set, the conv layer weights take on different dimensions, wrt to above example: (3,3,3,16) for 'tf' and (16,3,3,3) for 'th'.

A bit ugly, but code below should work for Keras 2:

p.s. Remember to set keras.backend.set_image_dim_ordering('th') for Keras 2 as well....
def load_weights(model,yolo_weight_file):           
    
    tiny_data = np.fromfile(yolo_weight_file,np.float32)[4:]
  
    index = 0
    for layer in model.layers:
        weights = layer.get_weights()
        if len(weights)>0:
            filter_shape, bias_shape = [w.shape for w in weights]
            if len(filter_shape)>2: #For convolutional layers
                filter_shape_i = filter_shape[::-1]
                bias_weight = tiny_data[index:index+np.prod(bias_shape)].reshape(bias_shape)
                index += np.prod(bias_shape)
                filter_weight= tiny_data[index:index+np.prod(filter_shape_i)].reshape(filter_shape_i)
                filter_weight= np.transpose(filter_weight,(2,3,1,0))
                index += np.prod(filter_shape)
                layer.set_weights([filter_weight,bias_weight])
            else: #For regular hidden layers
                bias_weight = tiny_data[index:index+np.prod(bias_shape)].reshape(bias_shape)
                index += np.prod(bias_shape)
                filter_weight= tiny_data[index:index+np.prod(filter_shape)].reshape(filter_shape)
                index += np.prod(filter_shape)
                layer.set_weights([filter_weight,bias_weight])

这才是大神

JoyceYa · 2018-11-08T13:38:59Z

@Alteregoxxx Hi, i had the same problem, did you find a solution?

miscla · 2018-12-28T09:34:39Z

@Alteregoxxx Hi, i had the same problem, did you find a solution?

did you find a solution??

Zhongan-Wang · 2019-03-10T09:18:24Z

Why i use this method but it did't work?
为什么我加上之后，也没有作用

pbvelikov · 2020-03-28T21:13:30Z

Hi All,
I have the same problem. Is there a solution for keras 2.2.4?

xslittlegrass closed this as completed Mar 27, 2017

xslittlegrass reopened this Apr 27, 2017

docPhil99 mentioned this issue Aug 31, 2017

Bounding boxes not showing subodh-malgonde/vehicle-detection#2

Closed

anandkoirala mentioned this issue Nov 17, 2017

No bounding boxes #11

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Incompatible with Keras > 2.0 #3

Incompatible with Keras > 2.0 #3

SaschaHornauer commented Mar 22, 2017

xslittlegrass commented Mar 27, 2017

im2ex commented Apr 27, 2017 •

edited

Loading

xslittlegrass commented Apr 27, 2017

liborw commented May 5, 2017

marc-chan commented Jul 27, 2017 •

edited

Loading

sungmin-yang commented Dec 6, 2017

XiongLianga commented Mar 29, 2018

Alteregoxxx commented Oct 9, 2018

Wentaobi commented Oct 11, 2018

JoyceYa commented Nov 8, 2018

miscla commented Dec 28, 2018

Zhongan-Wang commented Mar 10, 2019

pbvelikov commented Mar 28, 2020

Incompatible with Keras > 2.0 #3

Incompatible with Keras > 2.0 #3

Comments

SaschaHornauer commented Mar 22, 2017

xslittlegrass commented Mar 27, 2017

im2ex commented Apr 27, 2017 • edited Loading

xslittlegrass commented Apr 27, 2017

liborw commented May 5, 2017

marc-chan commented Jul 27, 2017 • edited Loading

sungmin-yang commented Dec 6, 2017

XiongLianga commented Mar 29, 2018

Alteregoxxx commented Oct 9, 2018

Wentaobi commented Oct 11, 2018

JoyceYa commented Nov 8, 2018

miscla commented Dec 28, 2018

Zhongan-Wang commented Mar 10, 2019

pbvelikov commented Mar 28, 2020

im2ex commented Apr 27, 2017 •

edited

Loading

marc-chan commented Jul 27, 2017 •

edited

Loading