Download
Reliable Fidelity and Diversity Metrics for Generative Models (ICML 2020). Devising indicative evaluation metrics for the image generation task remains an open problem. The most widely used metric for measuring the similarity between real and generated images has been the Fréchet Inception Distance (FID) score. Because it does not differentiate the fidelity and diversity aspects of the generated images, recent papers have introduced variants of precision and recall metrics to diagnose those properties separately. In this paper, we show that even the latest version of the precision and recall (Kynkäänniemi et al., 2019) metrics are not reliable yet. For example, they fail to detect the match between two identical distributions, they are not robust against outliers, and the evaluation hyperparameters are selected arbitrarily. We propose density and coverage metrics that solve the above issues.
Features
- Precision and recall metrics
- Density and coverage metrics
- Test 10000 real and fake samples form the standard normal distribution N(0,I) in 1000-dimensional Euclidean space
- Generating many fake samples around the real outlier is enough to increase the precision measure
- Set the nearest neighbour k=5
- Precision, recall, density, and coverage estimates
Project Samples
