QCC-WM: Linking Macro-State Dynamics and Visual Generation via Quotient Causal Construction
cs.CV
Submitted: 2026-11-09
Updated: 2026-11-09
Comments: Published in Medical Image Analysis
Journal ref: Medical Image Analysis (2026), 104391
DOI: 10.1016/j.media.2026.104391
Code: https://github.com/chengtan9907/OpenSTL
License: http://creativecommons.org/licenses/by/4.0/
The gist: A key challenge in action-conditioned world modeling is to construct macro-states from high-dimensional observations that support recursive prediction under actions…
Terminology
Abstract
A key challenge in action-conditioned world modeling is to construct macro-states from high-dimensional observations that support recursive prediction under actions, while keeping their predicted evolution consistent with visual changes in generated observations. We propose QCC-WM, a visual world model based on Quotient-based Causal Construction. QCC-WM uses action responses to guide macro-variable construction for recursive action-conditioned prediction. A shared reference quotient measurement constrains the consistency between predicted states and observed changes, allowing macro-level predictions to provide structured constraints for future visual generation. Using translational and rotational actions and their predicted responses, the model aggregates local visual features into macro-nodes with spatial correspondence through causal abstraction, and recursively predicts node content and motion under a shared reference measurement. Macro-level predictions guide future observation generation, while predicted motion transports historical appearance information. Node-level errors between generated observations and macro-level targets are then mapped back to the corresponding spatial regions for correction, improving consistency between visual generation and macro-level predictions. On the US-VLA dataset, QCC-WM achieves an ROI-MSE of 0.01861 for four-step prediction. Compared with the best baseline at each prediction horizon, it reduces LPIPS by 48.59%, 41.68%, and 30.23% for four-step prediction and eight- and sixteen-step recursive extrapolation, respectively. Ablation studies and mechanism analyses show that motion-guided reference transport mainly improves pixel reconstruction, while node-level error feedback further improves perceptual quality and consistency between generated observations and macro-level predictions.
Sources
- Slot-JEPA: Object-Structured Prediction Interfaces for Latent World Models
- Bridging the Gap to Real-World Object-Centric Learning
- FLIP: Flow-Centric Generative Planning as General-Purpose Manipulation World Model
- DeXplicit: Making Physical State Explicit in Dexterous Manipulation World Models
- PredRNN: A Recurrent Neural Network for Spatiotemporal Predictive Learning
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs