PreSIST: Vision-Language-Informed Object Persistence Prediction in Open-World Scenes

2026-07-04

AI Summary

This work addresses the need for long-term deployed robots to proactively predict the persistence of objects in dynamic environments, rather than passively updating beliefs only upon revisiting. The authors propose PreSIST, a novel framework that leverages vision-language models (VLMs) to jointly reason about object attributes and scene context, inferring human activity and usage patterns to generate instance-level persistence priors. To facilitate efficient real-world deployment, they further introduce PreSIST-Vis, a lightweight variant that operates without language input. The approach integrates pseudo-label training, probabilistic persistence filtering, and survival time modeling. Evaluated on a newly curated in-the-wild object persistence dataset, both PreSIST-Lang and PreSIST-Vis significantly outperform existing baselines.

Abstract

Robots deployed over long periods must reason about environments that change over time. Existing long-term perception systems often address object change reactively, updating their maps only after revisiting a scene and observing that an object has moved. Instead, robots should reason proactively about how long objects are likely to persist using the context in which they appear. For example, a car at a traffic light and a car in a parking spot share the same semantic class, but their contexts imply different persistence durations. We propose PreSIST (Predictive Scene-conditioned Instance Survival over Time), a method for predicting whether an observed object will remain in its last seen pose at arbitrary future times. PreSIST estimates instance-level persistence priors from object properties and scene context, then integrates these priors with a probabilistic persistence filter as observations become available. Its key insight is that the reasoning capabilities of vision-language models (VLMs) can relate scene context to likely object use and human activity, enabling persistence prediction before long-term observations are available. We develop two interchangeable variants: PreSIST-Lang, which estimates persistence priors using a VLM, and PreSIST-Vis, a novel vision-only model trained using PreSIST-Lang pseudo-labels for efficient deployment. Experiments on a new dataset of in-the-wild object persistence annotations show that PreSIST-Lang and PreSIST-Vis outperform baselines on open-world persistence prediction.

Problem

• object persistence prediction
• open-world scenes
• scene context
• long-term perception
• vision-language models

Innovation

• vision-language models
• object persistence prediction
• scene context reasoning
• probabilistic filtering
• open-world perception