Search papers, labs, and topics across Lattice.
The paper introduces Adjoint Schrödinger Bridge Matching (ASBM), a generative modeling framework that learns optimal data trajectories by first learning a Schrödinger Bridge (SB) forward dynamic to couple data to an energy-defined prior, and then learning the backward generative dynamic using a matching loss. By employing a non-memoryless forward process, ASBM achieves straighter sampling paths, leading to improved stability and efficiency in high-dimensional data generation. Experiments on image generation demonstrate that ASBM improves fidelity with fewer sampling steps and can be distilled into a one-step generator.
Ditch the noisy diffusion detour: ASBM finds a surprisingly direct route from data to noise, slashing sampling steps and boosting image fidelity.
Diffusion models often yield highly curved trajectories and noisy score targets due to an uninformative, memoryless forward process that induces independent data-noise coupling. We propose Adjoint Schrödinger Bridge Matching (ASBM), a generative modeling framework that recovers optimal trajectories in high dimensions via two stages. First, we view the Schrödinger Bridge (SB) forward dynamic as a coupling construction problem and learn it through a data-to-energy sampling perspective that transports data to an energy-defined prior. Then, we learn the backward generative dynamic with a simple matching loss supervised by the induced optimal coupling. By operating in a non-memoryless regime, ASBM produces significantly straighter and more efficient sampling paths. Compared to prior works, ASBM scales to high-dimensional data with notably improved stability and efficiency. Extensive experiments on image generation show that ASBM improves fidelity with fewer sampling steps. We further showcase the effectiveness of our optimal trajectory via distillation to a one-step generator.
