フロー反転による条件付きソース分布学習：時間的フローマッチングの新手法CNP-Flow
Learning Conditional Source Distribution via Flow Reversal for Temporal Flow Matching
条件付きフローマッチングにおいて、条件ごとにガウスソース分布を学習するCNP-Flowを提案し、動画予測・補間・ロボット動作計画で生成品質を向上させた。
著者: Kuan-Hsun Tu, Hsuan-Chi Liu, Jia-Wei Liao, Chien-Sheng Chiang, Tsung-Wei Ke
分類: cs.CV
原文アブストラクト
We introduce CNP-Flow, a flow matching framework for temporal generation that learns conditional source distributions through flow reversal. Whereas standard conditional flow matching (FM) incorporates conditioning through the vector field and draws source samples from a standard Gaussian, CNP-Flow uses a conditional noise predictor (CNP) to produce an isotropic Gaussian source for each temporal condition. The CNP is supervised by source samples obtained through flow reversal, which maps observed targets backward through a pretrained FM model. A three-stage pipeline pretrains the FM model, trains the CNP, and fine-tunes the FM model using the learned source distribution, while preserving the FM backbone architecture. Across video prediction, video interpolation, and 7-DoF Franka robot motion planning, CNP-Flow consistently improves generation quality. It also matches baseline performance with fewer function evaluations. Project page: https://embodiedai-ntu.github.io/cnpflow