New Study Proposes Learning Anisotropic Prediction Targets for Flow-Based Generative Models
An arXiv paper introduces an approach for learning anisotropic prediction targets in flow-based generative models, which are typically trained on hand-chosen targets such as velocity, noise, or clean data.
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-11-16 05:00 · arXiv cs.CV
Learning Anisotropic Prediction Targets for Flow Models