hhl134
hhl134
Impressive code! When running it with my own dataset, I utilized SuperPoint to extract feature points and proceeded with matching. However, I encountered an issue with too few feature points,...
在depths, masks = load_colmap(image_list, args.datadir,args.depth_H, args.depth_W)这个函数里 def load_point_vis(path, masks): with open(path, 'rb') as f: n = struct.unpack('
Have you tested the precision when used in conjunction with SuperGlue? I found that the results were not very good in my tests.