Particle filtering (PFs) and, more generally, sequential Monte Carlo (SMC) methods are essential tools for Bayesian inference. Over the years, many SMC variants have been proposed, yet their core always relies on importance sampling followed by a resampling step.While resampling is crucial to mitigate particle degeneracy and to maintain a stable approximation of the posterior distribution, it often represents a significant computational bottleneck.In this work, we present a novel, fast, resampling procedure that provides significant computational gains in demanding (often high-dimensional) scenarios where a large number of particles is required, and the effective sample size (ESS) is small.The effectiveness of the proposed approach is demonstrated through a series of numerical experiments showing remarkable performance. In addition, a theoretical analysis and related code implementation are provided.