Paper
Geração de Imagem
The FID Lottery: Quantifying Hidden Randomness in Generative-Model Evaluation
Analysis of FID variance across different training and sampling seeds reveals significant reproducibility issues in image generation evaluation, with retraining causing larger fluc…
Hugging Face · Daily Papers
·Nicolas Dufour, Alexei A. Efros
·
·▲ 6 upvotes
Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.
Autores: Nicolas Dufour, Alexei A. Efros, Patrick Pérez
- 6 upvotes da comunidade
- Temas: Frechet Inception Distance, FID, training seeds, generation seeds, classifier-free-guidance, flow-matching loss
Resumo
Resumo original (em inglês), extraído do paper:
Analysis of FID variance across different training and sampling seeds reveals significant reproducibility issues in image generation evaluation, with retraining causing larger fluctuations than resampling, and recommends updated evaluation protocols with error bars and optimal guidance tuning.
