MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Fu, Yuxiang, Yan, Qi, Wang, Lele, Li, Ke, Liao, Renjie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916651405934592
author Fu, Yuxiang
Yan, Qi
Wang, Lele
Li, Ke
Liao, Renjie
author_facet Fu, Yuxiang
Yan, Qi
Wang, Lele
Li, Ke
Liao, Renjie
contents In this paper, we address the problem of human trajectory forecasting, which aims to predict the inherently multi-modal future movements of humans based on their past trajectories and other contextual cues. We propose a novel motion prediction conditional flow matching model, termed MoFlow, to predict K-shot future trajectories for all agents in a given scene. We design a novel flow matching loss function that not only ensures at least one of the $K$ sets of future trajectories is accurate but also encourages all $K$ sets of future trajectories to be diverse and plausible. Furthermore, by leveraging the implicit maximum likelihood estimation (IMLE), we propose a novel distillation method for flow models that only requires samples from the teacher model. Extensive experiments on the real-world datasets, including SportVU NBA games, ETH-UCY, and SDD, demonstrate that both our teacher flow model and the IMLE-distilled student model achieve state-of-the-art performance. These models can generate diverse trajectories that are physically and socially plausible. Moreover, our one-step student model is $\textbf{100}$ times faster than the teacher flow model during sampling. The code, model, and data are available at our project page: https://moflow-imle.github.io
format Preprint
id arxiv_https___arxiv_org_abs_2503_09950
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation
Fu, Yuxiang
Yan, Qi
Wang, Lele
Li, Ke
Liao, Renjie
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
In this paper, we address the problem of human trajectory forecasting, which aims to predict the inherently multi-modal future movements of humans based on their past trajectories and other contextual cues. We propose a novel motion prediction conditional flow matching model, termed MoFlow, to predict K-shot future trajectories for all agents in a given scene. We design a novel flow matching loss function that not only ensures at least one of the $K$ sets of future trajectories is accurate but also encourages all $K$ sets of future trajectories to be diverse and plausible. Furthermore, by leveraging the implicit maximum likelihood estimation (IMLE), we propose a novel distillation method for flow models that only requires samples from the teacher model. Extensive experiments on the real-world datasets, including SportVU NBA games, ETH-UCY, and SDD, demonstrate that both our teacher flow model and the IMLE-distilled student model achieve state-of-the-art performance. These models can generate diverse trajectories that are physically and socially plausible. Moreover, our one-step student model is $\textbf{100}$ times faster than the teacher flow model during sampling. The code, model, and data are available at our project page: https://moflow-imle.github.io
title MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2503.09950