Distribution Preserving Multiple Hypotheses Prediction for Uncertainty Modeling

2021-10-06 · Tobias Leemann, Moritz Sackmann, Jörn Thielecke, Ulrich Hofmann

Many supervised machine learning tasks, such as future state prediction in dynamical systems, require precise modeling of a forecast's uncertainty. The Multiple Hypotheses Prediction (MHP) approach addresses this problem by providing several hypotheses that represent possible outcomes. Unfortunately, with the common $l_2$ loss function, these hypotheses do not preserve the data distribution's characteristics. We propose an alternative loss for distribution preserving MHP and review relevant theorems supporting our claims. Furthermore, we empirically show that our approach yields more representative hypotheses on a synthetic and a real-world motion prediction data set. The outputs of the proposed method can directly be used in sampling-based Monte-Carlo methods.

📄 PDF Abstract BibTeX arXiv:2110.02858

Code (1)

tleemann/dpmhp 공식 구현

Tasks

motion predictionPrediction

Hypothesis Disparity Regularized Mutual Information Maximization

2020-12-15 · Qicheng Lao, Xiang Jiang, Mohammad Havaei

We propose a hypothesis disparity regularized mutual information maximization~(HDMI) approach to tackle unsupervised hypothesis transfer -- as an effort towards unifying hypothesis transfer learning (HTL) and unsupervise…

Domain Adaptation Transfer Learning Unsupervised Domain Adaptation

Multi-person Pose Tracking using Sequential Monte Carlo with Probabilistic Neural Pose Predictor

2019-09-16 · Masashi Okada, Shinji Takenaka, Tadahiro Taniguchi

It is an effective strategy for the multi-person pose tracking task in videos to employ prediction and pose matching in a frame-by-frame manner. For this type of approach, uncertainty-aware modeling is essential because …

Pose Tracking Prediction

Solomonoff-Inspired Hypothesis Ranking with LLMs for Prediction Under Uncertainty

2025-12-19 · Josh Barber, Rourke Young, Cameron Coombe, Will Browne arxiv

Reasoning under uncertainty is a key challenge in AI, especially for real-world tasks, where problems with sparse data demands systematic generalisation. Existing approaches struggle to balance accuracy and simplicity wh…

SnapPose3D: Diffusion-Based Single-Frame 2D-to-3D Lifting of Human Poses

2026-04-29 · Alessandro Simoni, Riccardo Catalini, Davide Di Nucci, Guido Borghi 외 arxiv

Depth ambiguity and joint uncertainty are the two main obstacles in obtaining accurate human pose predictions by 2D-to-3D lifting methods proposed in the literature. In particular, these issues are caused by 2D joint loc…

3D Human Pose Estimation Temporal Sequences

Using Soft Labels to Model Uncertainty in Medical Image Segmentation

2021-09-26 · João Lourenço Silva, Arlindo L. Oliveira

Medical image segmentation is inherently uncertain. For a given image, there may be multiple plausible segmentation hypotheses, and physicians will often disagree on lesion and organ boundaries. To be suited to real-worl…

Image Segmentation Medical Image Segmentation Segmentation Semantic Segmentation+1