Struct Classifications 

pub struct Classifications {
    pub predictions: Vec<ClassPrediction>,
}

The classification results for a single image (or crop): one ClassPrediction per class the classifier scored.

Fields

predictions: Vec<ClassPrediction>

One entry per class, in whatever order the classifier produced them (typically class id order, not sorted by confidence). Use Classifications::top_k / Classifications::top1 for a confidence-sorted view.

Implementations

impl Classifications

pub fn new(predictions: Vec<ClassPrediction>) -> Self

Creates a new Classifications from an existing prediction list.

pub fn from_scores(scores: &[f32]) -> Self

Builds a Classifications from a dense per-class score vector (scores[class_id]), as produced directly by a softmax/sigmoid output layer.

pub fn top_k(&self, k: usize) -> Vec<ClassPrediction>

The k highest-confidence predictions, sorted descending. Returns fewer than k if there aren’t that many predictions.

pub fn top1(&self) -> Option<ClassPrediction>

The single highest-confidence prediction, if any.

Trait Implementations

impl Clone for Classifications

fn clone(&self) -> Classifications

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Classifications

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

Formats the value using the given formatter.

impl Default for Classifications

fn default() -> Classifications

Returns the “default value” for a type.

impl PartialEq for Classifications

fn eq(&self, other: &Classifications) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for Classifications