---
title: "Towards Robust Deterministic and Probabilistic Modeling for Predictive Learning"
authors:
  - "Xuesong Nie"
  - "Haoyuan Jin"
  - "Vijayakumar Bhagavatula"
  - "Xiaofeng Liu"
doi: "10.24963/ijcai.2025/195"
url: "https://ijcai.org/proceedings/2025/195"
pdf: "https://ijcai.org/proceedings/2025/0195.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "1747-1755"
track: "Main Track"
keywords:
  - "Computer Vision: CV: Video analysis and understanding"
  - "Computer Vision: CV: Motion and tracking"
  - "Machine Learning: ML: Self-supervised Learning"
---

# Towards Robust Deterministic and Probabilistic Modeling for Predictive Learning

Xuesong Nie, Haoyuan Jin, Vijayakumar Bhagavatula, Xiaofeng Liu

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 1747-1755. DOI: [10.24963/ijcai.2025/195](https://doi.org/10.24963/ijcai.2025/195)

## Abstract

Predictive modeling of unannotated spatiotemporal data presents inherent challenges, primarily due to the highly entangled visual dynamics in real-world scenes. To tackle these complexities, we introduce a novel insight through Disentangling Deterministic and Probabilistic (DDP) modeling. We note a key observation in spatiotemporal data where low-level details typically remain stable, whereas high-level motion frequently exhibits dynamic variations. The core motivation involves constructing two distinct pathways in the latent space: a deterministic path and a probabilistic path. The probabilistic path begins by defining the motion flow, which explicitly describes complex many-to-many motion patterns between patches, and models its probabilistic distribution using a motion diffuser. The deterministic path incorporates a spectral-aware enhancer to retain and amplify visual details in the frequency domain. These designs ensure visual consistency while also capturing intricate long-term motion dynamics. Extensive experiments demonstrate the superiority of DDP across diverse scenario evaluations.