Base class for pointpillars_nuscenes .
Input is points data and related params.
Output is a struct of detection results, named PointPillarsNuscenesResult
.
Sample code :
...
std::string anno_file_name = "./sample_pointpillars_nus.info";
PointsInfo points_info;
std::string model_0 = "pointpillars_nuscenes_40000_64_0_pt";
std::string model_1 = "pointpillars_nuscenes_40000_64_1_pt";
auto pointpillars = vitis::ai::PointPillarsNuscenes::create(
model_0, model_1);
auto points_dim = pointpillars->getPointsDim();
read_inno_file_pp_nus(anno_file_name, points_info, points_dim,
points_info.sweep_infos);
auto ret = pointpillars->run(points_info);
...
please see the test sample for detail.
Quick Function Reference
The following table lists all the functions defined in the vitis::ai::PointPillarsNuscenes
class:
Type | Name | Arguments |
---|---|---|
std::unique_ptr< PointPillarsNuscenes > |
create |
|
std::unique_ptr< PointPillarsNuscenes > |
create |
|
int | getInputWidth |
|
int | getInputHeight |
|
size_t | get_input_batch |
|
int | getPointsDim |
|
std::vector< float > | sweepsFusionFilter |
|
std::vector< std::vector< float > > | sweepsFusionFilter |
|
PointPillarsNuscenesResult
|
run |
|
std::vector< PointPillarsNuscenesResult > |
run |
|
PointPillarsNuscenesResult
|
run |
|
std::vector< PointPillarsNuscenesResult > |
run |
|