search

XuyangBai / D3Feat

[TensorFlow] Official implementation of CVPR'20 oral paper - D3Feat: Joint Learning of Dense Detection and Description of 3D Local Features https://arxiv.org/abs/2003.03164
MIT License
259 stars 38 forks source link

Why i cannot get the first dimension of the features' shape? #48

Closed BiaoBiaoLi closed 2 years ago

BiaoBiaoLi commented 2 years ago

Thanks your work! In the D3Feat.py, i print the shape,

def assemble_FCNN_blocks(inputs, config, dropout_prob):
    # First get features from CNN  
    F = assemble_CNN_blocks(inputs, config, dropout_prob)
    features = F[-1]
    for f in F:
        print("the feature shape ", f.shape)

but get the below: image but i see the dimension of the features in paper are : 64 128 256 512 1024

XuyangBai commented 2 years ago

it is the feature of TensorFlow. you can use tf.Print to printout the real value of features

BiaoBiaoLi commented 2 years ago

Thanks，i get it