Data Structure Index

The following is a list of data structures:

vitis::ai::ANNORET

Struct of the result returned by the pointpillars neuron network in the annotation mode. It is mainly used for accuracy test or bev image drawing.

Declaration

typedef struct
{
  std::vector< std::string > name,
  V1I label,
  V1F truncated,
  V1I occluded,
  V1F alpha,
  V2I bbox,
  V2F dimensions,
  V2F location,
  V1F rotation_y,
  V1F score,
  V2F box3d_camera,
  V2F box3d_lidar  void clear(),

} vitis::ai::ANNORET;
Table 1. Structure vitis::ai::ANNORET member description
MemberDescription
nameName of detected result in vector: such as Car Cylist Pedestrian.
labelLabel of detected result.
truncatedTruncated information.
occludedOccluded information.
alphaAlpha information.
bboxbbox information.
dimensionsDimensions information.
locationLocation information.
rotation_yrotation_y information.
scoreScore information.
box3d_camerabox3d_camera information.
box3d_lidarbox3d_lidar information.
clearInner function to clear all fields.

vitis::ai::BCCResult

Declaration

typedef struct
{
  int width,
  int height,
  int count
} vitis::ai::BCCResult;
Table 2. Structure vitis::ai::BCCResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
countCount of crowd.

vitis::ai::CenterPointResult

Struct of the result with the centerpoint network.

Declaration

typedef struct
{
  std::vector< float > bbox,
  float score,
  int label
} vitis::ai::CenterPointResult;
Table 3. Structure vitis::ai::CenterPointResult member description
MemberDescription
bboxBounding box 3d: {x, y, z, x_size, y_size, z_size, yaw}
scoreScore
labelClassification

vitis::ai::ClassificationResult

Struct of the result with the classification network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< Score > scores,
  int type  const char * lookup(int index),

} vitis::ai::ClassificationResult;
Table 4. Structure vitis::ai::ClassificationResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
scores

A vector of object width confidence in the first k; k defaults to 5 and can be modified through the model configuration file.

typeClassification label type.
lookupThe classification corresponding by index.

vitis::ai::ClassificationResult::Score

The struct of index and confidence for an object.

Declaration

typedef struct
{
  int index,
  float score
} vitis::ai::ClassificationResult::Score;
Table 5. Structure vitis::ai::ClassificationResult::Score member description
MemberDescription
indexThe index of the result in the ImageNet.
scoreConfidence of this category.

vitis::ai::Covid19SegmentationResult

Declaration

typedef struct
{
  int width,
  int height,
  cv::Mat positive_classification,
  cv::Mat infected_area_classification
} vitis::ai::Covid19SegmentationResult;
Table 6. Structure vitis::ai::Covid19SegmentationResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
positive_classificationPositive detection result. The cv::Mat type is CV_8UC1 or CV_8UC3.
infected_area_classificationInfected area detection result. The cv::Mat type is CV_8UC1 or CV_8UC3.

vitis::ai::DISPLAY_PARAM

Four data structure getting from the calibration information. It is mainly used for accuracy test or bev image drawing. See detail in the overview/samples/pointpillars/readme for more information.

Declaration

typedef struct
{
  V2F P2,
  V2F rect,
  V2F Trv2c,
  V2F p2rect
} vitis::ai::DISPLAY_PARAM;
Table 7. Structure vitis::ai::DISPLAY_PARAM member description
MemberDescription
P2P2 information.
rectrect information
Trv2cTrv2c information.
p2rectp2rect information

vitis::ai::FaceDetectResult

Struct of the result with the facedetect network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > rects
} vitis::ai::FaceDetectResult;
Table 8. Structure vitis::ai::FaceDetectResult member description
MemberDescription
widthWidth of an input image.
heightHeight of an input image.
rectsAll faces, filtered by confidence >= detect threshold.

vitis::ai::FaceDetectResult::BoundingBox

The coordinate and confidence of a face.

Declaration

typedef struct
{
  float x,
  float y,
  float width,
  float height,
  float score
} vitis::ai::FaceDetectResult::BoundingBox;
Table 9. Structure vitis::ai::FaceDetectResult::BoundingBox member description
MemberDescription
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

face width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

face height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

scoreface confidence, the value range from 0 to 1.

vitis::ai::FaceFeatureFixedResult

The result of FaceFeature. It is a 512 dimensions vector, fix point values.

Declaration

typedef struct
{
  std::array< int8_t, 512 > vector_t  int width,
  int height,
  float scale  std::unique_ptr< vector_t > feature,

} vitis::ai::FaceFeatureFixedResult;
Table 10. Structure vitis::ai::FaceFeatureFixedResult member description
MemberDescription
vector_tThe 512 dimensions vector, in fix point format.
widthWidth of an input image.
heightHeight of an input image.
scaleThe fix point.
featureA vector of 512 fixed values.

vitis::ai::FaceFeatureFloatResult

The result of FaceFeature. It is a 512 dimensions vector, float value.

Declaration

typedef struct
{
  std::array< float, 512 > vector_t  int width,
  int height  std::unique_ptr< vector_t > feature,

} vitis::ai::FaceFeatureFloatResult;
Table 11. Structure vitis::ai::FaceFeatureFloatResult member description
MemberDescription
vector_tThe 512 dimensions vector.
widthWidth of an input image.
heightHeight of an input image.
featureA vector of 512 float values.

vitis::ai::FaceLandmarkResult

Struct of the result returned by the facelandmark network.

Declaration

typedef struct
{
  std::array< std::pair< float, float >, 5 > points
} vitis::ai::FaceLandmarkResult;
Table 12. Structure vitis::ai::FaceLandmarkResult member description
MemberDescription
points

Five key points coordinate. This array of <x,y> has five elements, x / y is normalized relative to width / height, the value range from 0 to 1.

vitis::ai::FaceQuality5ptResult

The struct of result returned by the facequality5pt network.

Declaration

typedef struct
{
  int width,
  int height,
  float score,
  std::array< std::pair< float, float >, 5 > points
} vitis::ai::FaceQuality5ptResult;
Table 13. Structure vitis::ai::FaceQuality5ptResult member description
MemberDescription
widthWidth of a input image.
heightHeight of a input image.
score

The quality of face. The value range is from 0 to 1. If the option "original_quality" in the model prototxt is false, it is a normal mode. If the option "original_quality" is true, the quality score can be larger than 1, this is a special mode only for accuracy test.

points

Five key points coordinate. An array of <x,y> has five elements where x and y are normalized relative to input image columns and rows. The value range is from 0 to 1.

vitis::ai::HourglassResult

Result with the openpose network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< PosePoint > poses
} vitis::ai::HourglassResult;
Table 14. Structure vitis::ai::HourglassResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
poses

A vector of pose, pose is represented by a vector of PosePoint. Joint points are arranged in order 0: head, 1: neck, 2: L_shoulder, 3:L_elbow, 4: L_wrist, 5: R_shoulder, 6: R_elbow, 7: R_wrist, 8: L_hip, 9:L_knee, 10: L_ankle, 11: R_hip, 12: R_knee, 13: R_ankle

vitis::ai::HourglassResult::PosePoint

Struct of a coordinate point and the point type.

Declaration

typedef struct
{
  int type,
  cv::Point2f point
} vitis::ai::HourglassResult::PosePoint;
Table 15. Structure vitis::ai::HourglassResult::PosePoint member description
MemberDescription
typePoint type.
  • 1 : "valid"
  • 3 : "invalid"
pointCoordinate point.

vitis::ai::MedicalDetectionResult

Struct of the result returned by the medical refinedet network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes
} vitis::ai::MedicalDetectionResult;
Table 16. Structure vitis::ai::MedicalDetectionResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
bboxesAll objects, a vector of BoundingBox.

vitis::ai::MedicalDetectionResult::BoundingBox

Struct of an object coordinate ,confidence and classification.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height
} vitis::ai::MedicalDetectionResult::BoundingBox;
Table 17. Structure vitis::ai::MedicalDetectionResult::BoundingBox member description
MemberDescription
labelClassification.
scoreConfidence.
x

x-coordinate. x is normalized relative to the input image columns. The range of values is from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. The range of values is from 0 to 1.

width

width. width is normalized relative to the input image columns. The value range from 0 to 1.

height

height, height is normalized relative to the input image rows. The value range from 0 to 1.

vitis::ai::MedicalSegcellResult

Struct of the result returned by the segmentation neuron network.

Declaration

typedef struct
{
  int width,
  int height,
  cv::Mat segmentation
} vitis::ai::MedicalSegcellResult;
Table 18. Structure vitis::ai::MedicalSegcellResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
segmentationSegmentation result in cv::Mat mode.

vitis::ai::MedicalSegmentationResult

Struct of the result returned by the MedicalSegmentation neuron network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< cv::Mat > segmentation
} vitis::ai::MedicalSegmentationResult;
Table 19. Structure vitis::ai::MedicalSegmentationResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
segmentationA vector of cv::Mat (segmentation result).

vitis::ai::MultiTaskResult

Struct of the result returned by the MultiTask network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< VehicleResult > vehicle,
  cv::Mat segmentation
} vitis::ai::MultiTaskResult;
Table 20. Structure vitis::ai::MultiTaskResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
vehicleDetection result of SSD task.
segmentationSegmentation result to visualize, cv::Mat type is CV_8UC1 or CV_8UC3.

vitis::ai::MultiTaskv3Result

Struct of the result returned by the MultiTaskv3 network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< Vehiclev3Result > vehicle,
  cv::Mat segmentation,
  cv::Mat lane,
  cv::Mat drivable,
  cv::Mat depth
} vitis::ai::MultiTaskv3Result;
Table 21. Structure vitis::ai::MultiTaskv3Result member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
vehicleDetection result of SSD task.
segmentationSegmentation result to visualize, cv::Mat type is CV_8UC1 or CV_8UC3.
laneLane segmentation.
drivableDrivable area.
depthDepth estimation.

vitis::ai::OpenPoseResult

Result with the openpose network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< std::vector< PosePoint > > poses
} vitis::ai::OpenPoseResult;
Table 22. Structure vitis::ai::OpenPoseResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
poses

A vector of pose. Pose is represented by a vector of PosePoint. Joint points are arranged in order 0: head, 1: neck, 2: L_shoulder, 3:L_elbow, 4: L_wrist, 5: R_shoulder, 6: R_elbow, 7: R_wrist, 8: L_hip, 9:L_knee, 10: L_ankle, 11: R_hip, 12: R_knee, 13: R_ankle

vitis::ai::OpenPoseResult::PosePoint

Struct of a coordinate point and the point type.

Declaration

typedef struct
{
  int type,
  cv::Point2f point
} vitis::ai::OpenPoseResult::PosePoint;
Table 23. Structure vitis::ai::OpenPoseResult::PosePoint member description
MemberDescription
typePoint type.
  • 1 : "valid"
  • 3 : "invalid"
pointCoordinate point.

vitis::ai::PlateDetectResult

Struct of the result returned by the platedetect network.

Declaration

typedef struct
{
  int width,
  int height,
  BoundingBox box,
  Point top_left,
  Point top_right,
  Point bottom_left,
  Point bottom_right
} vitis::ai::PlateDetectResult;
Table 24. Structure vitis::ai::PlateDetectResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
boxThe position of plate.
top_leftThe top_left point.
top_rightThe top_right point.
bottom_leftThe bottom_left point.
bottom_rightThe bottom_right point.

vitis::ai::PlateDetectResult::BoundingBox

Declaration

typedef struct
{
  float score,
  float x,
  float y,
  float width,
  float height
} vitis::ai::PlateDetectResult::BoundingBox;
Table 25. Structure vitis::ai::PlateDetectResult::BoundingBox member description
MemberDescription
scorePlate confidence, the value ranges from 0 to 1.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Plate width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Plate height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

vitis::ai::PlateDetectResult::Point

Plate coordinate point.

Declaration

typedef struct
{
  float x,
  float y
} vitis::ai::PlateDetectResult::Point;
Table 26. Structure vitis::ai::PlateDetectResult::Point member description
MemberDescription
xx-coordinate, the value ranges from 0 to 1.
yy-coordinate, the value ranges from 0 to 1.

vitis::ai::PlateNumResult

Struct of the result of the platenum network.

Declaration

typedef struct
{
  int width,
  int height,
  std::string plate_number,
  std::string plate_color
} vitis::ai::PlateNumResult;
Table 27. Structure vitis::ai::PlateNumResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
plate_numberThe plate number.
plate_colorThe plate color, Blue / Yellow.

vitis::ai::pointpillars_nus::CamInfo

Camera information of input points.

Declaration

typedef struct
{
  uint64_t timestamp,
  std::array< float, 3 > s2l_t,
  std::array< float, 9 > s2l_r,
  std::array< float, 9 > cam_intr
} vitis::ai::pointpillars_nus::CamInfo;
Table 28. Structure vitis::ai::pointpillars_nus::CamInfo member description
MemberDescription
timestampTimestamp of input points.
s2l_tSensor to lidar translation params.
s2l_rSensor to lidar rotation params.
cam_intrCamera intrinsic params.

vitis::ai::pointpillars_nus::Points

Structure to describe input points data.

Declaration

typedef struct
{
  int dim,
  std::shared_ptr< std::vector< float > > points
} vitis::ai::pointpillars_nus::Points;
Table 29. Structure vitis::ai::pointpillars_nus::Points member description
MemberDescription
dimPoints dim.
pointsPoints data.

vitis::ai::pointpillars_nus::PointsInfo

Structure to describe points information.

Declaration

typedef struct
{
  std::vector< CamInfo > cam_info,
  Points points,
  uint64_t timestamp,
  std::vector< SweepInfo > sweep_infos
} vitis::ai::pointpillars_nus::PointsInfo;
Table 30. Structure vitis::ai::pointpillars_nus::PointsInfo member description
MemberDescription
cam_infoCamera information.
pointsPoints.
timestampTimestamp of points.
sweep_infosSweeps information.

vitis::ai::pointpillars_nus::SweepInfo

Structure to describe sweeps.

Declaration

typedef struct
{
  CamInfo cam_info,
  Points points
} vitis::ai::pointpillars_nus::SweepInfo;
Table 31. Structure vitis::ai::pointpillars_nus::SweepInfo member description
MemberDescription
cam_infoCamera information.
pointsPoints.

vitis::ai::PointPillarsNuscenesResult

Struct of the result returned by the PointPillarsNuscenes network.

Declaration

typedef struct
{
  std::vector< PPBbox > bboxes
} vitis::ai::PointPillarsNuscenesResult;
Table 32. Structure vitis::ai::PointPillarsNuscenesResult member description
MemberDescription
bboxesAll bounding boxes.

vitis::ai::PointPillarsResult

Struct of the final result returned by the pointpillars neuron network encapsulated with width/height information.

Declaration

typedef struct
{
  int width,
  int height,
  PPResult ppresult
} vitis::ai::PointPillarsResult;
Table 33. Structure vitis::ai::PointPillarsResult member description
MemberDescription
widthWidth of network input.
heightHeight of network input.
ppresultFinal result returned by the pointpillars neuron network.

vitis::ai::PoseDetectResult

Struct of the result returned by the posedetect network.

Declaration

typedef struct
{
  cv::Point2f Point  int width,
  int height  Pose14Pt pose14pt,

} vitis::ai::PoseDetectResult;
Table 34. Structure vitis::ai::PoseDetectResult member description
MemberDescription
PointA coordinate points.
widthWidth of input image.
heightHeight of input image.
pose14ptThe pose of input image.

vitis::ai::PoseDetectResult::Pose14Pt

A pose. Represented by 14 coordinate points.

Declaration

typedef struct
{
  Point right_shoulder,
  Point right_elbow,
  Point right_wrist,
  Point left_shoulder,
  Point left_elbow,
  Point left_wrist,
  Point right_hip,
  Point right_knee,
  Point right_ankle,
  Point left_hip,
  Point left_knee,
  Point left_ankle,
  Point head,
  Point neck
} vitis::ai::PoseDetectResult::Pose14Pt;
Table 35. Structure vitis::ai::PoseDetectResult::Pose14Pt member description
MemberDescription
right_shoulderR_shoulder coordinate.
right_elbowR_elbow coordinate.
right_wristR_wrist coordinate.
left_shoulderL_shoulder coordinate.
left_elbowL_elbow coordinate.
left_wristL_wrist coordinate.
right_hipR_hip coordinate.
right_kneeR_knee coordinate.
right_ankleR_ankle coordinate.
left_hipL_hip coordinate.
left_kneeL_knee coordinate.
left_ankleL_ankle coordinate.
headHead coordinate.
neckNeck coordinate.

vitis::ai::PPBbox

Struct of an object coordinate, confidence and classification. .

Declaration

typedef struct
{
  float score,
  std::vector< float > bbox,
  uint32_t label
} vitis::ai::PPBbox;
Table 36. Structure vitis::ai::PPBbox member description
MemberDescription
scoreConfidence.
bbox

Bounding box: x, y, z, x-size, y-size, z-size, yaw, custom value and so on.

labelClassification.

vitis::ai::PPResult

Struct of the final result returned by the pointpillars neuron network.

Declaration

typedef struct
{
  V2F final_box_preds,
  V1F final_scores,
  V1I label_preds
} vitis::ai::PPResult;
Table 37. Structure vitis::ai::PPResult member description
MemberDescription
final_box_predsFinal box predicted.
final_scoresFinal scores predicted.
label_predsFinal label predicted.

vitis::ai::RcanResult

Result with the Rcan network.

Declaration

typedef struct
{
  int width,
  int height,
  cv::Mat feat
} vitis::ai::RcanResult;
Table 38. Structure vitis::ai::RcanResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
featDouble size of input image.

vitis::ai::RefineDetResult

Struct of the result with the refinedet network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes
} vitis::ai::RefineDetResult;
Table 39. Structure vitis::ai::RefineDetResult member description
MemberDescription
widthWidth of the input image.
heightHeight of the input image.
bboxesThe vector of BoundingBox.

vitis::ai::RefineDetResult::BoundingBox

Struct of an object coordinates and confidence.

Declaration

typedef struct
{
  float x,
  float y,
  float width,
  float height,
  int label,
  float score
} vitis::ai::RefineDetResult::BoundingBox;
Table 40. Structure vitis::ai::RefineDetResult::BoundingBox member description
MemberDescription
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Body width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Body height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

labelBody detection label. The value ranges from 0 to 21.
scoreBody detection confidence. The value ranges from 0 to 1.

vitis::ai::ReidResult

Result with the ReID network.

Declaration

typedef struct
{
  int width,
  int height,
  cv::Mat feat
} vitis::ai::ReidResult;
Table 41. Structure vitis::ai::ReidResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
featThe feature of input image.

vitis::ai::RetinaFaceResult

Struct of the result with the retinaface network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes,
  std::vector< std::array< std::pair< float, float >, 5 > > landmarks
} vitis::ai::RetinaFaceResult;
Table 42. Structure vitis::ai::RetinaFaceResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
bboxesAll faces, filtered by confidence >= detect threshold.
landmarksLandmarks.

vitis::ai::RetinaFaceResult::BoundingBox

The coordinate and confidence of a face.

Declaration

typedef struct
{
  float x,
  float y,
  float width,
  float height,
  float score
} vitis::ai::RetinaFaceResult::BoundingBox;
Table 43. Structure vitis::ai::RetinaFaceResult::BoundingBox member description
MemberDescription
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Face width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Face height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

scoreFace confidence. The value ranges from 0 to 1.

vitis::ai::RoadLineResult

Struct of the result returned by the roadline network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< Line > lines
} vitis::ai::RoadLineResult;
Table 44. Structure vitis::ai::RoadLineResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
linesThe vector of line.

vitis::ai::RoadLineResult::Line

Struct of the result returned by the roadline network.

Declaration

typedef struct
{
  int type,
  std::vector< cv::Point > points_cluster
} vitis::ai::RoadLineResult::Line;
Table 45. Structure vitis::ai::RoadLineResult::Line member description
MemberDescription
type

Road line type, the value range from 0 to 3.

  • 0 : background
  • 1 : white dotted line
  • 2 : white solid line
  • 3 : yollow line
points_clusterPoint clusters, make line from these.

vitis::ai::Segmentation3DResult

Base class for segmentation 3D object data in the vector<float> mode.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< int > array
} vitis::ai::Segmentation3DResult;
Table 46. Structure vitis::ai::Segmentation3DResult member description
MemberDescription
widthWidth of the network model.
heightHeight of the network model.
arrayInput 3D object data.

vitis::ai::SegmentationResult

Struct of the result returned by the segmentation network.

FPN Num of segmentation classes

  • 0 : "unlabeled"
  • 1 : "ego vehicle"
  • 2 : "rectification border"
  • 3 : "out of roi"
  • 4 : "static"
  • 5 : "dynamic"
  • 6 : "ground"
  • 7 : "road"
  • 8 : "sidewalk"
  • 9 : "parking"
  • 10 : "rail track"
  • 11 : "building"
  • 12 : "wall"
  • 13 : "fence"
  • 14 : "guard rail"
  • 15 : "bridge"
  • 16 : "tunnel"
  • 17 : "pole"
  • 18 : "polegroup"

Declaration

typedef struct
{
  int width,
  int height,
  cv::Mat segmentation
} vitis::ai::SegmentationResult;
Table 47. Structure vitis::ai::SegmentationResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
segmentationSegmentation result. The cv::Mat type is CV_8UC1 or CV_8UC3.

vitis::ai::SSDResult

Struct of the result returned by the SSD neuron network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes
} vitis::ai::SSDResult;
Table 48. Structure vitis::ai::SSDResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
bboxesAll objects, a vector of BoundingBox.

vitis::ai::SSDResult::BoundingBox

Struct of an object coordinate, confidence and classification.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height
} vitis::ai::SSDResult::BoundingBox;
Table 49. Structure vitis::ai::SSDResult::BoundingBox member description
MemberDescription
labelClassification.
scoreConfidence.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

vitis::ai::TFSSDResult

Struct of the result returned by the TFSSD neuron network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes
} vitis::ai::TFSSDResult;
Table 50. Structure vitis::ai::TFSSDResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
bboxesAll objects, a vector of BoundingBox.

vitis::ai::TFSSDResult::BoundingBox

Struct of an object coordinate, confidence, classification.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height
} vitis::ai::TFSSDResult::BoundingBox;
Table 51. Structure vitis::ai::TFSSDResult::BoundingBox member description
MemberDescription
labelClassification.
scoreConfidence.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

vitis::ai::VehicleResult

A struct to define detection result of MultiTask.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height,
  float angle
} vitis::ai::VehicleResult;
Table 52. Structure vitis::ai::VehicleResult member description
MemberDescription
label

number of classes

  • 0 : "background"
  • 1 : "person"
  • 2 : "car"
  • 3 : "truck"
  • 4 : "bus"
  • 5 : "bike"
  • 6 : "sign"
  • 7 : "light"
scoreConfidence of this target.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

angleThe angle between the target vehicle and ourself.

vitis::ai::Vehiclev3Result

A struct to define detection result of MultiTaskv3.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height,
  float angle
} vitis::ai::Vehiclev3Result;
Table 53. Structure vitis::ai::Vehiclev3Result member description
MemberDescription
label

number of classes

  • 0 : "car"
  • 1 : "sign"
  • 2 : "person"
scoreConfidence of this target.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

angleThe angle between the target vehicle and ourself.

vitis::ai::YOLOv2Result

Struct of the result returned by the YOLOv2 network.

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes
} vitis::ai::YOLOv2Result;
Table 54. Structure vitis::ai::YOLOv2Result member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
bboxesAll objects.

vitis::ai::YOLOv2Result::BoundingBox

Struct of an object coordinate, confidence, and classification.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height
} vitis::ai::YOLOv2Result::BoundingBox;
Table 55. Structure vitis::ai::YOLOv2Result::BoundingBox member description
MemberDescription
labelClassification.
scoreConfidence. The value ranges from 0 to 1.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

vitis::ai::YOLOv3Result

Struct of the result returned by the YOLOv3 neuron network.
Note: VOC dataset category:string label[20] = {"aeroplane", "bicycle", "bird", "boat", "bottle", "bus","car", "cat", "chair", "cow", "diningtable", "dog", "horse", "motorbike","person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"};
Note: ADAS dataset category : string label[3] = {"car", "person", "cycle"};

Declaration

typedef struct
{
  int width,
  int height,
  std::vector< BoundingBox > bboxes
} vitis::ai::YOLOv3Result;
Table 56. Structure vitis::ai::YOLOv3Result member description
MemberDescription
widthWidth of input image.
heightHeight of output image.
bboxesAll objects, The vector of BoundingBox.

vitis::ai::YOLOv3Result::BoundingBox

Struct of detection result with an object.

Declaration

typedef struct
{
  int label,
  float score,
  float x,
  float y,
  float width,
  float height
} vitis::ai::YOLOv3Result::BoundingBox;
Table 57. Structure vitis::ai::YOLOv3Result::BoundingBox member description
MemberDescription
labelClassification.
scoreConfidence. The value ranges from 0 to 1.
x

x-coordinate. x is normalized relative to the input image columns. Range from 0 to 1.

y

y-coordinate. y is normalized relative to the input image rows. Range from 0 to 1.

width

Width. Width is normalized relative to the input image columns, Range from 0 to 1.

height

Height. Heigth is normalized relative to the input image rows, Range from 0 to 1.

xilinx::ai::FaceQualityResult

The result of the facequality network. It is a single float value.

Declaration

typedef struct
{
  int width,
  int height,
  float value
} xilinx::ai::FaceQualityResult;
Table 58. Structure xilinx::ai::FaceQualityResult member description
MemberDescription
widthWidth of input image.
heightHeight of input image.
valueQuality value ranges from 0.0 to 1.0.