TypeError: super(type, obj): obj must be an instance or subtype of type

[kumarshrestha009]

TypeError

in () 1 import UnariesNet ----> 2 uNet = UnariesNet.unariesNet() /home/kumar/DeepOcclusion/UnariesNet.py in __init__(self, load_pretrained) 80 c53_r = mNet.c53_r 81 ---> 82 op = ROIPoolingOp(pooled_h=p_h, pooled_w=p_w, spatial_scale=1.0) 83 84 /home/kumar/DeepOcclusion/roi_pooling.py in __init__(self, pooled_h, pooled_w, spatial_scale) 18 19 def __init__(self, pooled_h, pooled_w, spatial_scale): ---> 20 super(ROIPoolingOp, self).__init__(self.func_file, 21 self.func_name) 22 TypeError: super(type, obj): obj must be an instance or subtype of type How can i fix this ? Help will be highly appreciated ?

[DhanasekarChellamuthu]

Hi @kumarshrestha009, before that in same roi_pooling.py 7 th line from theano.sandbox.cuda import as_cuda_ndarray_variable, GpuOp i am getting error like cannot import name 'as_cuda_ndarray_variable' #38. can you help me to resolve this issue I am having doubt at Dependencies 2 nd point:

You also need to install the theano ROI-Pooling Layer from https://github.com/ddtm/theano-roi-pooling to get the unary potentials to work.

can you comment system dependencies

[kumarshrestha009]

You can type import theano.sandbox.cuda as cuda_ndarray_variable

Theano is the main dependency. I had used theano 0.9.

[DhanasekarChellamuthu]

Thanks for answering @kumarshrestha009 but you said like calling whole package in the name as_cuda_ndarray_variable but in that code it is written as calling one function from that package that both make different meaning right.

[Kirmani66]

@kumarshrestha009 Did you solved the super(type, obj) issue. If yes please share the solution

[leconghieu99]

@DhanasekarChellamuthu can you please send me the .zip files. Looks like the link is broken. Email: Hieus.lecong@gmai.com