Crate voirs_evaluation

Expand description

§`VoiRS` Evaluation

Comprehensive quality evaluation and assessment framework for the VoiRS ecosystem. This crate provides tools for evaluating speech synthesis quality, pronunciation accuracy, and comparative analysis between different models or systems.

§Features

Quality Evaluation: Objective and subjective quality metrics
Pronunciation Assessment: Phoneme-level accuracy scoring
Comparative Analysis: Side-by-side evaluation of different systems
Perceptual Metrics: Human-perception-aligned quality measures
Automated Scoring: ML-based quality prediction

§Quick Start

use voirs_evaluation::quality::QualityEvaluator;
use voirs_evaluation::traits::QualityEvaluator as QualityEvaluatorTrait;
use voirs_sdk::AudioBuffer;

    // Create quality evaluator
    let evaluator = QualityEvaluator::new().await?;
     
    // Create test audio buffers
    let generated = AudioBuffer::new(vec![0.1; 16000], 16000, 1);
    let reference = AudioBuffer::new(vec![0.12; 16000], 16000, 1);
     
    // Evaluate quality
    let quality = evaluator.evaluate_quality(&generated, Some(&reference), None).await?;
    println!("Quality score: {:.2}", quality.overall_score);

Re-exports§

pub use performance::multi_gpu;
pub use performance::LRUCache;
pub use performance::PersistentCache;
pub use performance::SlidingWindowProcessor;
pub use traits::*;

Modules§

accuracy_benchmarks: Comprehensive accuracy benchmarking for speech synthesis systems.
advanced_preprocessing: Advanced Audio Preprocessing for Evaluation Enhancement
audio: Comprehensive audio I/O and format support for VoiRS evaluation.
audit: Audit trail system for compliance and security monitoring Audit trail system for compliance, debugging, and security monitoring
automated_benchmarks: Automated Benchmark Update System
benchmark_export: Benchmark result export and import functionality
benchmark_runner: Automated benchmark runner with regression detection.
benchmarks: Benchmark tests against reference implementations
caching: Commercial tool comparison framework for speech evaluation systems Advanced result caching system with multiple backends Advanced Result Caching System
commercial_tool_comparison: Commercial tool comparison framework for speech evaluation systems
comparison: Comparative evaluation implementation
compliance: Industry standard compliance module
compliance_testing: Compliance testing suite for standards validation Compliance Testing Suite for Standards Validation
context_aware: Context-aware evaluation system for speech synthesis Context-aware evaluation system for speech synthesis quality assessment.
conversational: Conversational quality assessment for dialogue systems Conversational quality assessment for dialogue and interactive speech systems.
cpp_bindings: C++ header-only interface for evaluation framework C++ header-only interface for VoiRS evaluation framework
cross_language_validation: Cross-language evaluation accuracy validation framework Cross-language evaluation accuracy validation framework
csf_validation: Critical Success Factors (CSF) validation framework Critical Success Factors (CSF) Validation Framework
data_quality_validation: Advanced data quality validation and dataset management utilities Advanced data quality validation and dataset management utilities
data_versioning: Data versioning system for benchmark and evaluation results Data Versioning System for Benchmark and Evaluation Results
dataset_management: Evaluation Dataset Management System
deep_learning_metrics: Deep learning-based evaluation metrics with neural MOS prediction Deep Learning-Based Evaluation Metrics
distributed: Distributed evaluation system for parallel processing across multiple nodes
doc_generation: Automated documentation generation for evaluation results Automated Documentation Generation for Evaluation Results
enterprise_security: Enterprise security framework with RBAC, encryption, and compliance Enterprise Security Framework for VoiRS Evaluation
error_enhancement: Enhanced error message generation utilities
fairness: Fairness-aware evaluation for bias detection and demographic parity
federated: Federated evaluation system for distributed processing Federated Evaluation System
fuzzing: Fuzzing tests for robustness validation
graphql: GraphQL API for complex evaluation queries GraphQL API for Complex Evaluation Queries
ground_truth_dataset: Ground truth dataset management for evaluation validation Ground truth dataset management for evaluation validation
integration: VoiRS Ecosystem Integration
kubernetes: Kubernetes deployment configuration for distributed evaluation Kubernetes Deployment Configuration for Distributed Evaluation
logging: Enhanced logging and debugging utilities Enhanced logging and debugging support for VoiRS evaluation
matlab_bindings: MATLAB/Octave bindings for evaluation framework MATLAB/Octave bindings for VoiRS evaluation framework
metric_reliability_testing: Metric reliability and reproducibility testing framework Metric reliability and reproducibility testing framework
metrics_comparison: Metrics comparison and regression detection Metrics comparison and regression detection
metrics_explainability: Metrics explainability system for interpretable evaluation
multi_turn_dialogue: Multi-turn dialogue evaluation for extended conversations Multi-turn dialogue evaluation for extended conversations and dialogue systems.
multiregion: Multi-region deployment configuration for global distribution Multi-Region Deployment Configuration for Global Distribution
nodejs_bindings: Node.js/JavaScript bindings for evaluation framework Node.js/JavaScript bindings for VoiRS evaluation framework
observability: Monitoring and observability framework with Prometheus and tracing Monitoring and Observability Framework for VoiRS Evaluation
perceptual: Perceptual evaluation and listening test simulation
performance: Performance optimization utilities
performance_enhancements: Performance enhancement utilities for faster evaluation Performance Enhancement Module
performance_monitor: Advanced Performance Monitoring for VoiRS Evaluation
platform: Platform Compatibility Module
plugins: Plugin system for custom evaluation metrics Plugin system for custom evaluation metrics
precision: Numerical precision utilities for high-accuracy calculations
prelude: Convenient prelude for common imports Prelude module for convenient imports
privacy: Privacy-preserving evaluation framework with differential privacy Privacy-Preserving Evaluation Framework
pronunciation: Auto-generated module structure
protocol_documentation: Protocol documentation and compliance validation utilities Evaluation protocol documentation and compliance validation
quality: Quality evaluation implementations
quality_gates: Quality gate validation system for automated quality assurance Quality gate validation system for automated quality assurance
rbac: Role-Based Access Control (RBAC) for enterprise security Role-Based Access Control (RBAC) system for enterprise evaluation security
regression_detector: Automated regression detection for performance benchmarks.
regression_testing: Regression Testing Framework for VoiRS Evaluation
reproducibility: Reproducibility guarantees system for deterministic evaluation Reproducibility Guarantees System
rest_api: REST API interface for evaluation services REST API interface for evaluation services
semantic_similarity: Semantic similarity evaluation for speech content analysis Semantic similarity evaluation for speech synthesis systems.
standards: Industry standards compliance module (ANSI, ISO/IEC, AES, ITU-T) Industry Standards Compliance Module
statistical: Statistical analysis modules for evaluation metrics
statistical_enhancements: Enhanced statistical analysis utilities Statistical Analysis Enhancements
task_oriented: Task-oriented dialogue evaluation for goal-driven interactions
traits: Core traits for the VoiRS evaluation system
user_experience: User experience (UX) evaluation for usability and satisfaction
validation: Comprehensive Validation Framework
validation_certificates: Validation certificate generator for certified evaluation results
websocket: WebSocket interface for real-time evaluation services WebSocket interface for real-time evaluation services
workflows: Evaluation workflow system for automated pipelines Evaluation Workflow System

Macros§

eval_log: Macro for easy logging
profile_function: Macro for easy function profiling

Structs§

AudioBuffer: Audio buffer containing synthesized speech
Phoneme: Phoneme representation with IPA symbol and metadata
PhonemeAlignment: Phoneme alignment result
Transcript: Represents a transcribed text with metadata

Enums§

EvaluationError: Evaluation-specific error types
LanguageCode: Language code identifier
VoirsError: Main error type for VoiRS operations with enhanced categorization

Constants§

VERSION: Version information

Functions§

calculate_correlation: Calculate statistical correlation between two score vectors
default_comparison_config: Create a default comparison configuration
default_pronunciation_config: Create a default pronunciation evaluation configuration
default_quality_config: Create a default quality evaluation configuration
normalize_scores: Utility function to normalize scores to 0-1 range
pronunciation_score_to_label: Convert pronunciation score to human-readable label
quality_score_to_label: Convert quality score to human-readable label
validate_audio_compatibility: Validate audio compatibility for evaluation
weighted_average: Calculate weighted average of scores

Crate voirs_evaluation

Crate voirs_evaluation Copy item path

§VoiRS Evaluation

§Features

§Quick Start

Re-exports§

Modules§

Macros§

Structs§

Enums§

Constants§

Functions§

Crate voirs_evaluation

§`VoiRS` Evaluation