input DATA
[[[[[  0.   1.   2.   3.   4.]
    [  5.   6.   7.   8.   9.]
    [ 10.  11.  12.  13.  14.]
    [ 15.  16.  17.  18.  19.]]

   [[ 20.  21.  22.  23.  24.]
    [ 25.  26.  27.  28.  29.]
    [ 30.  31.  32.  33.  34.]
    [ 35.  36.  37.  38.  39.]]

   [[ 40.  41.  42.  43.  44.]
    [ 45.  46.  47.  48.  49.]
    [ 50.  51.  52.  53.  54.]
    [ 55.  56.  57.  58.  59.]]]


  [[[ 60.  61.  62.  63.  64.]
    [ 65.  66.  67.  68.  69.]
    [ 70.  71.  72.  73.  74.]
    [ 75.  76.  77.  78.  79.]]

   [[ 80.  81.  82.  83.  84.]
    [ 85.  86.  87.  88.  89.]
    [ 90.  91.  92.  93.  94.]
    [ 95.  96.  97.  98.  99.]]

   [[100. 101. 102. 103. 104.]
    [105. 106. 107. 108. 109.]
    [110. 111. 112. 113. 114.]
    [115. 116. 117. 118. 119.]]]]]
ONNX result
[[[[30. 31. 32. 33. 34.]
   [35. 36. 37. 38. 39.]
   [40. 41. 42. 43. 44.]
   [45. 46. 47. 48. 49.]]

  [[50. 51. 52. 53. 54.]
   [55. 56. 57. 58. 59.]
   [60. 61. 62. 63. 64.]
   [65. 66. 67. 68. 69.]]

  [[70. 71. 72. 73. 74.]
   [75. 76. 77. 78. 79.]
   [80. 81. 82. 83. 84.]
   [85. 86. 87. 88. 89.]]]]
Writting model
Set opencv input DATA
Opencv result
[[[[30. 31. 32. 33. 34.]
   [35. 36. 37. 38. 39.]
   [40. 41. 42. 43. 44.]
   [45. 46. 47. 48. 49.]]

  [[50. 51. 52. 53. 54.]
   [55. 56. 57. 58. 59.]
   [60. 61. 62. 63. 64.]
   [65. 66. 67. 68. 69.]]

  [[50. 51. 52. 53. 54.]
   [55. 56. 57. 58. 59.]
   [60. 61. 62. 63. 64.]
   [65. 66. 67. 68. 69.]]]]
OPENCV result
Quadratic error for node  :  133.33333
Max error for  node :  400.0

import numpy as np
import onnx
import onnxsim
import onnx.reference
import cv2 as cv

shape_ini = (4, 5)
N = 2
data = np.arange(0, 1 * N * 3 *shape_ini[0]* shape_ini[1]).reshape((1, N, 3, shape_ini[0], shape_ini[1])).astype(dtype=np.float32)    
shape_ini = data.shape 
select_axes = np.array([1], dtype=np.int64)
keepdims = 0
print("input DATA ")
print(data)
onnx_name = "testReduceMean"
out1 = onnx.helper.make_tensor_value_info('out1', onnx.TensorProto.FLOAT, [None, None, None, None, None])
inp0 = onnx.helper.make_tensor_value_info('inp0', onnx.TensorProto.FLOAT, [None, None, None, None, None])
node1 = onnx.helper.make_node( "ReduceMean", inputs=["inp0"], outputs=["out1"], keepdims=keepdims , axes=select_axes)
graph = onnx.helper.make_graph([node1], onnx_name,  [inp0], [out1])
onnx_model = onnx.helper.make_model(graph)

feeds = {'inp0': data}
sess = onnx.reference.ReferenceEvaluator(onnx_model)
res_onnx = sess.run(["out1"], feeds)
print("ONNX result")
print(res_onnx[0])
print("Writting model")
with open(onnx_name + ".onnx", "wb") as f:
    f.write(onnx_model.SerializeToString())

net = cv.dnn.readNet(onnx_name+'.onnx')
net.enableWinograd(False)
print("Set opencv input DATA ")

net.setInput(data)
res_ocv = net.forward()
print("Opencv result")
print(res_ocv)
print("Quadratic error for node  : ", np.mean((res_ocv -  res_onnx[0])**2))
print("Max error for  node : ", np.max((res_ocv -  res_onnx[0])**2))