IRIS: A Real-World Benchmark for Inverse Recovery and Identification of Physical Dynamic Systems from Monocular Video

2026-03-17

AI Summary

Existing unsupervised methods for physical parameter estimation from video lack a unified real-world benchmark and standardized evaluation protocol. To address this gap, this work proposes IRIS—the first high-fidelity experimental benchmark encompassing both single- and multi-body dynamical systems—comprising 220 videos captured at 4K resolution and 60 fps, accompanied by ground-truth physical parameters, governing equations, and uncertainty annotations. We introduce a standardized evaluation protocol that assesses parameter accuracy, identifiability, extrapolation capability, robustness, and equation selection. By integrating four complementary equation discovery strategies into a multi-step physics-informed loss framework, we establish strong baselines on IRIS and uncover systematic failure modes of current approaches. The dataset, annotations, evaluation toolkit, and code are fully open-sourced.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Intelligent Robots: State Estimation
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Abstract

Unsupervised physical parameter estimation from video lacks a common benchmark: existing methods evaluate on non-overlapping synthetic data, the sole real-world dataset is restricted to single-body systems, and no established protocol addresses governing-equation identification. This work introduces IRIS, a high-fidelity benchmark comprising 220 real-world videos captured at 4K resolution and 60\,fps, spanning both single- and multi-body dynamics with independently measured ground-truth parameters and uncertainty estimates. Each dynamical system is recorded under controlled laboratory conditions and paired with its governing equations, enabling principled evaluation. A standardized evaluation protocol is defined encompassing parameter accuracy, identifiability, extrapolation, robustness, and governing-equation selection. Multiple baselines are evaluated, including a multi-step physics loss formulation and four complementary equation-identification strategies (VLM temporal reasoning, describe-then-classify prompting, CNN-based classification, and path-based labelling), establishing reference performance across all IRIS scenarios and exposing systematic failure modes that motivate future research. The dataset, annotations, evaluation toolkit, and all baseline implementations are publicly released.

Problem

• inverse recovery
• physical parameter estimation
• governing-equation identification
• real-world benchmark
• monocular video

Innovation

• real-world benchmark
• inverse recovery
• governing-equation identification
• unsupervised physical parameter estimation
• multi-body dynamics