Det3D:GT-AUG实现细节

实现类：CenterPoint/det3d/core/sampler/sample_ops.py: DataBaseSamplerV2

build_dbsampler参数：

type="GT-AUG",enable=False,
db_info_path="data/nuScenes/dbinfos_train_10sweeps_withvelo.pkl",
sample_groups=[
    dict(car=2),
    dict(truck=3),
    dict(construction_vehicle=7),
    dict(bus=4),
    dict(trailer=6),
    dict(barrier=2),
    dict(motorcycle=6),
    dict(bicycle=6),
    dict(pedestrian=2),
    dict(traffic_cone=2),
],
db_prep_steps=[
    dict(
        filter_by_min_num_points=dict(
            car=5,
            truck=5,
            bus=5,
            trailer=5,
            construction_vehicle=5,
            traffic_cone=5,
            barrier=5,
            motorcycle=5,
            bicycle=5,
            pedestrian=5,
        )
    ),
    dict(filter_by_difficulty=[-1],),
],
global_random_rotation_range_per_object=[0, 0],
rate=1.0,

对应DataBaseSamplerV2构造函数

db_infos = pickle.load(cfg.db_info_path)
groups = cfg.sample_groups
db_prepor = [build_db_preprocess(c, logger=logger) for c in cfg.db_prep_steps]
rate = cfg.rate
grot_range = list(grot_range)
self._group_db_infos = db_infos
self._rate = rate
self._groups = groups
self._sample_classes = ['car', 'truck',...,'traffic_cone']
self._sample_max_nums = [2, 3,...,2]
for k, v in self._group_db_infos.items():
    self._sampler_dict[k] = prep.BatchSampler(v, k) # 此处v为单个类别的对象的db数据，k为类名，BatchSampler为一个用于批量Shuffle采样的类
self._global_rot_range = global_rot_range
self._enable_global_rot = True if np.abs(global_rot_range[0] - global_rot_range[1]) >= 1e-3 # 所以没有允许这个，为啥？

具体采样实现为sample_all，实际用到的参数有root_path, gt_boxes, gt_names, num_point_features

Pipeline

1st: 确定每个cls的采样数目：

for class_name, max_sample_num in zip(
    self._sample_classes, self._sample_max_nums
):
    sampled_num = int(
        max_sample_num - np.sum([n == class_name for n in gt_names])
    )
    sampled_num = np.round(self._rate * sampled_num).astype(np.int64)
    sampled_num_dict[class_name] = sampled_num
    sample_num_per_class.append(sampled_num)

2nd: 根据确定的采样数目进行采样，调用sample_class_v2

sampled记录采样信息，sampled_gt_boxes记录采样到的bbox，avoid_coll_boxes是已经采样到的对象的bbox和gt的bbox的累加

3rd: 从sampled中取出sample的点云和bbox，并返回相应的res

Sample_class_v2

Pipeline:

1、计算gt的bv box
2、计算采样得到的sp_boxes
3、获得valid_mask Array[bool: len(gt_boxes) + len(sp_boxes)]
4、如果允许global_rot，prep.noise_per_object_v3_ # place samples to any place in a circle
5、计算采样的bv box
6、根据总体的bv box，进行碰撞测试
7、返回无碰撞的对象