HuguesTHOMAS / KPConv

Hey,

You seem to use a wrong aggregation file.

Lines 206 to 233 in 07861d8

    
           vertex_data, faces = read_ply(join(path, scene, scene + '_vh_clean_2.ply'), triangular_mesh=True) 
        
           vertices = np.vstack((vertex_data['x'], vertex_data['y'], vertex_data['z'])).T 
        
           vertices_colors = np.vstack((vertex_data['red'], vertex_data['green'], vertex_data['blue'])).T 
        
           vertices_labels = np.zeros(vertices.shape[0], dtype=np.int32) 
        
           if new_path == self.train_path: 
        
               # Load alignment matrix to realign points 
        
               align_mat = None 
        
               with open(join(path, scene, scene + '.txt'), 'r') as txtfile: 
        
                   lines = txtfile.readlines() 
        
               for line in lines: 
        
                   line = line.split() 
        
                   if line[0] == 'axisAlignment': 
        
                       align_mat = np.array([float(x) for x in line[2:]]).reshape([4, 4]).astype(np.float32) 
        
               R = align_mat[:3, :3] 
        
               T = align_mat[:3, 3] 
        
               vertices = vertices.dot(R.T) + T 
        
               # Get objects segmentations 
        
               with open(join(path, scene, scene + '_vh_clean_2.0.010000.segs.json'), 'r') as f: 
        
                   segmentations = json.load(f) 
        
               segIndices = np.array(segmentations['segIndices']) 
        
               # Get objects classes 
        
               with open(join(path, scene, scene + '_vh_clean.aggregation.json'), 'r') as f: 
        
                   aggregation = json.load(f)

However, in the ScanNet repo https://github.com/ScanNet/ScanNet, it says:
|-- _vh_clean_2.ply
Cleaned and decimated mesh for semantic annotations
|-- .aggregation.json, _vh_clean.aggregation.json
Aggregated instance-level semantic annotations on lo-res, hi-res meshes, respectively
|-- _vh_clean_2.0.010000.segs.json, _vh_clean.segs.json
Over-segmentation of lo-res, hi-res meshes, respectively (referenced by aggregated semantic annotations)

Please have a check!

Best

You seem to be right but I have not used the Scannet dataset for a very long time now. I will make a correction and if there is any other problems, don't hesitate to post another issue.

Thanks for spotting this.

Best,
Hugues

You seem to be right but I have not used the Scannet dataset for a very long time now. I will make a correction and if there is any other problems, don't hesitate to post another issue.

Thanks for spotting this.

Best,
Hugues

Thank you for your quick reply! By the way, Could I ask about reason about creating a rasterize_mesh (finer point clouds) with very number of points for ScanNet training ? Why don't we just use the raw point cloud for training?

Best,
Zhengdi

	vertex_data, faces = read_ply(join(path, scene, scene + '_vh_clean_2.ply'), triangular_mesh=True)
	vertices = np.vstack((vertex_data['x'], vertex_data['y'], vertex_data['z'])).T
	vertices_colors = np.vstack((vertex_data['red'], vertex_data['green'], vertex_data['blue'])).T

	vertices_labels = np.zeros(vertices.shape[0], dtype=np.int32)
	if new_path == self.train_path:

	# Load alignment matrix to realign points
	align_mat = None
	with open(join(path, scene, scene + '.txt'), 'r') as txtfile:
	lines = txtfile.readlines()
	for line in lines:
	line = line.split()
	if line[0] == 'axisAlignment':
	align_mat = np.array([float(x) for x in line[2:]]).reshape([4, 4]).astype(np.float32)
	R = align_mat[:3, :3]
	T = align_mat[:3, 3]
	vertices = vertices.dot(R.T) + T

	# Get objects segmentations
	with open(join(path, scene, scene + '_vh_clean_2.0.010000.segs.json'), 'r') as f:
	segmentations = json.load(f)

	segIndices = np.array(segmentations['segIndices'])

	# Get objects classes
	with open(join(path, scene, scene + '_vh_clean.aggregation.json'), 'r') as f:
	aggregation = json.load(f)

Bug on Scannet.py