RoNFA: Robust Neural Field-based Approach for Few-Shot Image Classification with Noisy Labels

Name: RoNFA: Robust Neural Field-based Approach for Few-Shot Image Classification with Noisy Labels
Uploaded: 2026-10-22T03:23:24.351809+00:00
Duration: 467 s
Description: RoNFA is a model for few-shot image classification that handles noisy labels effectively. It uses a neural field approach with two fields—one for features and one for categories—and soft K-means clustering to create robust class prototypes. This method ensures high accuracy even when the support set contains incorrect labels, outperforming standard methods.

RoNFA is a model for few-shot image classification that handles noisy labels effectively. It uses a neural field approach with two fields—one for features and one for categories—and soft K-means clustering to create robust class prototypes. This method ensures high accuracy even when the support set contains incorrect labels, outperforming standard methods.

Open the full episode