Struct Detection 

pub struct Detection {
    pub id: DetectionId,
    pub class_id: u32,
    pub confidence: f32,
    pub bbox: BBox,
    pub embedding: Option<Vec<f32>>,
    pub metadata: TypedMetadata,
}

A single object detection within one frame.

Detections are produced by detection stages and stored in DetectionSet. They carry spatial, classification, and optional re-identification data.

Use Detection::builder() for ergonomic construction with optional fields.

Fields

id: DetectionId

Unique ID within this frame’s detection set.

class_id: u32

Numeric class identifier (model-defined).

confidence: f32

Detection confidence score, typically in [0, 1].

bbox: BBox

Axis-aligned bounding box in normalized [0, 1] coordinates.

embedding: Option<Vec<f32>>

Optional re-identification or feature embedding vector.

metadata: TypedMetadata

Extensible metadata (domain-specific fields, additional scores, etc.).

Implementations

impl Detection

pub fn builder( id: DetectionId, class_id: u32, confidence: f32, bbox: BBox, ) -> DetectionBuilder

Create a builder for a Detection.

Example

use nv_core::{BBox, DetectionId};
use nv_perception::Detection;

let det = Detection::builder(
    DetectionId::new(1),
    0,
    0.95,
    BBox::new(0.1, 0.2, 0.3, 0.4),
)
.embedding(vec![0.1, 0.2, 0.3])
.build();

Trait Implementations

impl Clone for Detection

fn clone(&self) -> Detection

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for Detection

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.