Inception Score

Module: generative_models.core.evaluation.metrics.image.inception_score

Source: generative_models/core/evaluation/metrics/image/inception_score.py

Overview

Inception Score implementation using JAX and NNX.

The Inception Score (IS) is a metric for evaluating generative models, particularly GANs, based on two criteria:

1. Quality: Generated images should contain clear, recognizable objects
2. Diversity: The model should generate diverse images across classes

Classes

InceptionScore

class InceptionScore

Functions

init

def __init__()

calculate_kl_divergence

def calculate_kl_divergence()

calculate_score

def calculate_score()

classifier

def classifier()

compute

def compute()

get_predictions

def get_predictions()

Module Statistics

• Classes: 1
• Functions: 6
• Imports: 5