some annos have '0' in size

[zjr-bit]

Hi, thanks for your great work. When I try to use V2XSim in mmdetection3D, I find that there are some zeros in key size in some sample_annotations. So I use the below code to find them:

from nuscenes.nuscenes import NuScenes
version = 'v2.0'
root_path = 'xxx'
nusc = NuScenes(version=version, dataroot=root_path, verbose=True)

'''impletment of nusc.list_categories()'''
import numpy as np
print('Category stats for split %s:' % nusc.version)

# Add all annotations.
categories = dict()
count = 0
has_zero = dict()
for record in nusc.sample_annotation:
    if record['category_name'] not in categories:
        categories[record['category_name']] = []
    if np.all(np.array(record['size'])>0):
        categories[record['category_name']].append(record['size'] + [record['size'][1] / record['size'][0]])
    else:
        # print('{} category_name: {}'.format(record['token'], record['category_name']))
        count += 1
        if record['category_name'] not in has_zero:
            has_zero[record['category_name']] = []
        else:
            has_zero[record['category_name']].append(record['token'])
            
# Print stats.
for name, stats in sorted(categories.items()):
    stats = np.array(stats)
    print('{:27} n={:5}, width={:5.2f}\u00B1{:.2f}, len={:5.2f}\u00B1{:.2f}, height={:5.2f}\u00B1{:.2f}, '
            'lw_aspect={:5.2f}\u00B1{:.2f}'.format(name[:27], stats.shape[0],
                                                    np.mean(stats[:, 0]), np.std(stats[:, 0]),
                                                    np.mean(stats[:, 1]), np.std(stats[:, 1]),
                                                    np.mean(stats[:, 2]), np.std(stats[:, 2]),
                                                    np.mean(stats[:, 3]), np.std(stats[:, 3])))
print('total num of sample_anns which has 0 size: {}'.format(count))
for name, toks in sorted(has_zero.items()):
    print('{}: {}'.format(name, len(toks)))

The output are as below:

Category stats for split v2.0:
vehicle.bicycle             n=163636, width= 0.86±0.00, len= 1.49±0.00, height= 1.08±0.00, lw_aspect= 1.73±0.00
vehicle.car                 n=3131521, width= 2.02±0.24, len= 4.47±0.82, height= 1.65±0.29, lw_aspect= 2.20±0.26
vehicle.emergency.police    n=159584, width= 2.04±0.00, len= 4.97±0.00, height= 1.55±0.00, lw_aspect= 2.44±0.00
vehicle.motorcycle          n=469049, width= 0.81±0.04, len= 2.21±0.13, height= 1.23±0.06, lw_aspect= 2.91±11.38
total num of sample_anns which has 0 size: 366383
vehicle.bicycle: 366382

I was wondering if this was an error or if there was a deliberate intention behind it?