Style Metrics

Module: benchmarks.metrics.style_metrics

Source: benchmarks/metrics/style_metrics.py

Overview

StyleGAN-Specific Metrics for Image Generation Evaluation.

This module provides comprehensive metrics for evaluating StyleGAN3 performance, including traditional image quality metrics and StyleGAN-specific evaluations.

Key Metrics:

• Fréchet Inception Distance (FID)
• Inception Score (IS)
• Learned Perceptual Image Patch Similarity (LPIPS)
• Style Mixing Quality
• Few-shot Adaptation Performance
• Translation and Rotation Equivariance

Classes

EquivarianceMetric

class EquivarianceMetric

FIDMetric

class FIDMetric

FewShotAdaptationMetric

class FewShotAdaptationMetric

InceptionV3Feature

class InceptionV3Feature

LPIPSMetric

class LPIPSMetric

PerceptualNet

class PerceptualNet

StyleGANMetrics

class StyleGANMetrics

StyleMixingMetric

class StyleMixingMetric

Functions

call

def __call__()

call

def __call__()

init

def __init__()

init

def __init__()

init

def __init__()

init

def __init__()

init

def __init__()

init

def __init__()

init

def __init__()

init

def __init__()

compute_all_metrics

def compute_all_metrics()

compute_distance

def compute_distance()

compute_fid

def compute_fid()

compute_statistics

def compute_statistics()

compute_style_mixing_quality

def compute_style_mixing_quality()

evaluate_adaptation

def evaluate_adaptation()

evaluate_equivariance

def evaluate_equivariance()

Module Statistics

• Classes: 8
• Functions: 17
• Imports: 4