Papers › Diverse Human Motion Prediction via Gumbel-Softmax Sampling from an Auxiliary Space

Diverse Human Motion Prediction via Gumbel-Softmax Sampling from an Auxiliary Space

15 Jul 2022arXiv:2207.07351archive 2025-07-28

Lingwei Dang, Yongwei Nie, Chengjiang Long, Qing Zhang, Guiqing Li

Diverse human motion prediction aims at predicting multiple possible future pose sequences from a sequence of observed poses. Previous approaches usually employ deep generative networks to model the conditional distribution of data, and then randomly sample outcomes from the distribution. While different results can be obtained, they are usually the most likely ones which are not diverse enough. Recent work explicitly learns multiple modes of the conditional distribution via a deterministic network, which however can only cover a fixed number of modes within a limited range. In this paper, we propose a novel sampling strategy for sampling very diverse results from an imbalanced multimodal distribution learned by a deep generative model. Our method works by generating an auxiliary space and smartly making randomly sampling from the auxiliary space equivalent to the diverse sampling from the target distribution. We propose a simple yet effective network architecture that implements this novel sampling strategy, which incorporates a Gumbel-Softmax coefficient matrix sampling method and an aggressive diversity promoting hinge loss function. Extensive experiments demonstrate that our method significantly improves both the diversity and accuracy of the samplings compared with previous state-of-the-art sampling approaches. Code and pre-trained models are available at https://github.com/Droliven/diverse_sampling.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

droliven/diverse_sampling officialmentioned in papermentioned on GitHubpytorch report
Droliven/DHMP_jittor officialmentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DiversityHuman Pose ForecastingHuman motion predictionmotion prediction

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Human Pose Forecasting Human3.6M DiverseSampling ADE 370 #24 of 33 Archive leaderboard report
Human Pose Forecasting Human3.6M DiverseSampling APD 15310 #24 of 33 Archive leaderboard report
Human Pose Forecasting Human3.6M DiverseSampling CMD 11.692 #24 of 33 Archive leaderboard report
Human Pose Forecasting Human3.6M DiverseSampling FDE 485 #24 of 33 Archive leaderboard report
Human Pose Forecasting Human3.6M DiverseSampling FID 2.083 #24 of 33 Archive leaderboard report
Human Pose Forecasting Human3.6M DiverseSampling MMADE 475 #24 of 33 Archive leaderboard report
Human Pose Forecasting Human3.6M DiverseSampling MMFDE 516 #24 of 33 Archive leaderboard report
Human Pose Forecasting HumanEva-I DHMP ADE@2000ms 220 #2 of 11 Archive leaderboard report
Human Pose Forecasting HumanEva-I DHMP APD@2000ms 6109 #2 of 11 Archive leaderboard report
Human Pose Forecasting HumanEva-I DHMP FDE@2000ms 234 #2 of 11 Archive leaderboard report
Human Pose Forecasting HumanEva-I DHMP MMADE@2000ms 342 #2 of 11 Archive leaderboard report
Human Pose Forecasting HumanEva-I DHMP MMFDE@2000ms 316 #2 of 11 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections