Conditional Flow Matching

How it works

Comparisons

Highlighted claims

• Conditional flow matching defines a probability path between Gaussian noise and high-resolution precipitation sampled from the data distribution. — Conditional Flow Matching for Probabilistic Downscaling of Maximum 3-day Snowfall in Alaska
• The learned velocity network minimizes mean squared error between predicted and target velocities while masking missing precipitation pixels. — Conditional Flow Matching for Probabilistic Downscaling of Maximum 3-day Snowfall in Alaska
• Sampling uses the trained velocity network as an ODE right-hand side from Gaussian noise to the generated field. — Conditional Flow Matching for Probabilistic Downscaling of Maximum 3-day Snowfall in Alaska
• Conditional flow matching is presented as a fast alternative to diffusion models because it samples through ODE integration. — Conditional Flow Matching for Probabilistic Downscaling of Maximum 3-day Snowfall in Alaska