Leopard Speech-to-Text
Rust API

API Reference for the Leopard Rust SDK (crates.io).

leopard::Leopard

pub struct Leopard

Struct for the Leopard Speech-to-Text engine.

leopard::Leopard.process()

pub fn process(&self, pcm: &[i16]) -> Result<LeopardTranscript, LeopardError>

Processes a given audio data and returns its transcription. The audio needs to have a sample rate equal to .sample_rate and be 16-bit linearly-encoded. This function operates on single-channel audio. If you wish to process data in a different sample rate or format consider using .process_file().

Parameters

• pcm &[i16] : Audio data.

Returns

• Result<LeopardTranscript, LeopardError> : If Ok, returns the inferred transcription as a LeopardTranscript object. If Err, a LeopardError object describing the error that was encountered.

leopard::Leopard.process_file()

pub fn process_file<P: AsRef<Path>>(&self, audio_path: P) -> Result<LeopardTranscript, LeopardError>

Processes a given audio file and returns its transcription. The supported audio file formats are: 3gp (AMR), FLAC, MP3, MP4/m4a (AAC), Ogg, WAV and WebM.

Parameters

• audio_path P : Absolute path to the audio file.

Returns

• Result<LeopardTranscript, LeopardError> : If Ok, returns the inferred transcription as a LeopardTranscript object. If Err, a LeopardError object describing the error that was encountered.

leopard::Leopard.sample_rate()

pub fn sample_rate(&self) -> u32

Gets the sample rate.

Returns

• u32 : Audio sample rate accepted by Leopard.

leopard::Leopard.version()

pub fn version(&self) -> &str

Gets the Leopard version.

Returns

• &str : Leopard version string.

leopard::LeopardBuilder

pub struct LeopardBuilder

Builder struct for creating an instance of Leopard.

leopard::LeopardBuilder.new()

pub fn new()) -> Self

Creates a new LeopardBuilder.

Returns

• LeopardBuilder : A new Builder object.

leopard::LeopardBuilder.access_key()

pub fn access_key<S: Into<String>>(&mut self, access_key: S) -> &mut Self

Modifies the access_key of a LeopardBuilder object.

Parameters

• access_key S : AccessKey obtained from Picovoice Console.

Returns

• LeopardBuilder : The modified Builder object.

leopard::LeopardBuilder.model_path()

pub fn model_path<P: Into<PathBuf>>(&mut self, model_path: P) -> &mut Self

Modifies the model_path of a LeopardBuilder object.

Parameters

• model_path P : Path to the file containing model parameters (.pv).

Returns

• LeopardBuilder : The modified Builder object.

leopard::LeopardBuilder.library_path()

pub fn library_path<P: Into<PathBuf>>(&mut self, library_path: P) -> &mut Self

Modifies the library_path of a LeopardBuilder object.

Parameters

• library_path P : Path to the Leopard library file.

Returns

• LeopardBuilder : The modified Builder object.

leopard::LeopardBuilder.enable_automatic_punctuation()

pub fn enable_automatic_punctuation(&mut self, enable_automatic_punctuation: bool) -> &mut Self

Modifies the enable_automatic_punctuation of a LeopardBuilder object.

Parameters

• enable_automatic_punctuation bool : Set to true to enable automatic punctuation insertion.

Returns

• LeopardBuilder : The modified Builder object.

leopard::LeopardBuilder.enable_diarization()

pub fn enable_diarization(&mut self, enable_diarization: bool) -> &mut Self

Modifies the enable_diarization of a LeopardBuilder object.

Parameters

• enable_automatic_punctuation bool : Set to true to enable speaker diarization, which allows Leopard to differentiate speakers as part of the transcription process. Word metadata will include a speakerTag to identify unique speakers.

Returns

• LeopardBuilder : The modified Builder object.

leopard::LeopardBuilder.init()

pub fn init(&self) -> Result<Leopard, LeopardError>

Creates an instance of Leopard from the LeopardBuilder object.

Returns

• Result<Leopard, LeopardError> : If Ok, an initialized instance of Leopard. If Err, an instance of LeopardError detailing the error that was encountered.

leopard::LeopardWord

#[derive(PartialEq, Clone, Debug)]
pub struct LeopardWord {
    pub word: String,
    pub start_sec: f32,
    pub end_sec: f32,
    pub confidence: f32,
    pub speaker_tag: i32,
}

Struct containing a transcribed word and associated word metadata.

• word String : Transcribed word.
• start_sec f32 : Start of word in seconds.
• end_sec f32 : End of word in seconds.
• confidence f32 : Transcription confidence. It is a number within [0, 1].
• speaker_tag i32 : The speaker tag is -1 if diarization is not enabled during initialization; otherwise, it's a non-negative integer identifying unique speakers, with 0 reserved for unknown speakers.

leopard::LeopardTranscript

#[derive(PartialEq, Clone, Debug)]
pub struct LeopardTranscript {
    pub transcript: String,
    pub words: Vec<LeopardWord>,
}

Struct containing a transcript and transcribed words from Leopard.

leopard::LeopardError

pub struct LeopardError

An error type describing any errors encountered by the Leopard SDK.

leopard::LeopardError.new()

pub fn new(status: LeopardErrorStatus, message: impl Into<String>) -> Self

Creates a new instance of LeopardError

Parameters

• status LeopardErrorStatus : Type of error encountered.
• message impl Into<String> : Message detailing the cause of the error.

leopard::LeopardErrorStatus

pub enum LeopardErrorStatus {
    LibraryError(PvStatus),
    LibraryLoadError,
    FrameLengthError,
    ArgumentError,
}

Enum describing possible error types returned from the Leopard Rust SDK.

leopard::PvStatus

pub enum PvStatus {
    SUCCESS,
    OUT_OF_MEMORY,
    IO_ERROR,
    INVALID_ARGUMENT,
    STOP_ITERATION,
    KEY_ERROR,
    INVALID_STATE,
    RUNTIME_ERROR,
    ACTIVATION_ERROR,
    ACTIVATION_LIMIT_REACHED,
    ACTIVATION_THROTTLED,
    ACTIVATION_REFUSED,
}

Error codes returned from the Leopard library.