Struct CTCLoss 

pub struct CTCLoss { /* private fields */ }

Connectionist Temporal Classification (CTC) Loss.

Used for sequence-to-sequence tasks where alignment is unknown. Common in speech recognition and OCR.

Reference: Graves et al., “Connectionist Temporal Classification” (2006)

Implementations

impl CTCLoss

pub fn new(blank_idx: usize) -> Self

Create CTC loss with specified blank token index.

pub fn blank_idx(&self) -> usize

pub fn forward( &self, log_probs: &[Vec<f32>], targets: &[usize], input_length: usize, target_length: usize, ) -> f32

Compute CTC loss using forward-backward algorithm.

Arguments

• log_probs - Log probabilities [T, C] (time x classes)
• targets - Target sequence (class indices, no blanks)
• input_length - Length of input sequence
• target_length - Length of target sequence

Trait Implementations

impl Clone for CTCLoss

fn clone(&self) -> CTCLoss

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for CTCLoss

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

Formats the value using the given formatter.