I found we can generate new data for training this network through Pixel2Mesh-master/GenerateData. I download the ShapeNetCore.v1 dataset and then use the scripts 1_sample_points.txt and 2_generate_normal.py to generate the point cloud of model.obj in ShapeNet.
But there is some problem that I try to run the third scripts (3_camera_transform.py), it needs input rendering_metadata.txt. However, in original Shapenet dataset, there is no file name rendering_metadata.txt.
How can I get the rendering_metadata.txt? or I just need to copy the exsit TXT flie to all the directory?
For example: copy the exsit flie "Pixel2Mesh/GenerateData/1a0bc9ab92c915167ae33d942430658c/rendering/rendering_metadata.txt" to my own directory "02876657/109d55a137c042f5760315ac3bf2c13e/"
I found we can generate new data for training this network through Pixel2Mesh-master/GenerateData. I download the ShapeNetCore.v1 dataset and then use the scripts 1_sample_points.txt and 2_generate_normal.py to generate the point cloud of model.obj in ShapeNet. But there is some problem that I try to run the third scripts (3_camera_transform.py), it needs input rendering_metadata.txt. However, in original Shapenet dataset, there is no file name rendering_metadata.txt. How can I get the rendering_metadata.txt? or I just need to copy the exsit TXT flie to all the directory? For example: copy the exsit flie "Pixel2Mesh/GenerateData/1a0bc9ab92c915167ae33d942430658c/rendering/rendering_metadata.txt" to my own directory "02876657/109d55a137c042f5760315ac3bf2c13e/"