Pontuação FID
A Fréchet Início Distance (FID) Score is a metric used to evaluate the quality of images generated by inteligência artificial models, particularly generative adversarial networks (GANs). It quantifies how similar the generated images are to real images from a dataset, providing a numerical score that reflects the fidelity and diversity of the generated images.
Para calcular o FID Score, uma rede pré-treinada rede neural convolucional (CNN), often the Inception v3 model, is used to extract feature representations from both real and generated images. The key steps in calculating the FID Score involve:
- Extração de Características: As imagens são passadas pela CNN para obter vetores de características de alto nível.
- Análise Estatística: The mean and covariance of these feature vectors are computed for both real and generated image sets.
- Cálculo de Distância: The FID Score is then calculated using the Fréchet distance between the two Gaussiana multivariada distribuições definidas por essas estatísticas.
A lower FID Score indicates that the generated images are closer to the real images, suggesting higher quality. Conversely, a higher score implies that the generated images are less similar to the real ones. The FID Score is particularly useful because it takes into account both the quality (fidelity) and the diversity of the generated images, making it more reliable than simpler metrics como comparação pixel a pixel.
Em resumo, o FID Score serve como um importante parâmetro de referência no campo de IA geração de imagens, helping researchers and practitioners assess and improve their models.