Talk by Alexandre Alouadi, BNP Global Markets, CIFRE PhD
Title : LightSBB-M: Bridging Schrodinger and Bass for Generative Modeling
Abstract: The Schrodinger Bridge and Bass (SBB) formulation, which jointly controls drift and volatility, is an established extension of the classical Schrodinger Bridge (SB). Building on this framework, we introduce LightSBB-M, an algorithm that computes the optimal SBB transport plan in only a few iterations. The method exploits a dual representation of the SBB objective to obtain analytic expressions for the optimal drift and volatility, and it incorporates a tunable parameter β > 0 that interpolates between pure drift (the Schrodinger Bridge) and pure volatility (Bass ¨ martingale transport). We show that LightSBB-M achieves the lowest 2-Wasserstein distance on synthetic datasets against state-of-the-art Schrodinger ¨ Bridge and diffusion baselines with up to 35% improvement. We also illustrate the generative capability of the framework on an unpaired image-to-image translation task (adult ↔ child faces in FFHQ). These findings demonstrate that LightSBB-M provides a scalable, high-fidelity SBB solver that outperforms existing Schrodinger Bridge and diffusion baselines across both synthetic and real-world generative tasks. Moreover, we apply our algorithm for time series data.