EgoCross: Benchmarking Multimodal Large Language Models for Cross-Domain Egocentric Video Question Answering

TL;DR

EgoCross is a new benchmark designed to evaluate how well multimodal large language models can generalize across diverse real-world egocentric video domains like surgery, sports, and animals, revealing current models' limitations.

Contribution

The paper introduces EgoCross, a comprehensive cross-domain benchmark for egocentric video question answering, covering four challenging domains and providing a platform for evaluating and improving model generalization.

Findings

Most existing MLLMs struggle to generalize beyond daily activities.

Fine-tuning and reinforcement learning show potential for improvement.

EgoCross highlights significant gaps in current model robustness.

Abstract

Recent advances in Multimodal Large Language Models (MLLMs) have significantly pushed the frontier of egocentric video question answering (EgocentricQA). However, existing benchmarks and studies are mainly limited to common daily activities such as cooking and cleaning. In contrast, real-world deployment inevitably encounters domain shifts, where target domains differ substantially in both visual style and semantic content. To bridge this gap, we introduce \textbf{EgoCross}, a comprehensive benchmark designed to evaluate the cross-domain generalization of MLLMs in EgocentricQA. EgoCross covers four diverse and challenging domains, including surgery, industry, extreme sports, and animal perspective, representing realistic and high-impact application scenarios. It comprises approximately 1,000 QA pairs across 798 video clips, spanning four key QA tasks: prediction, recognition, localization, and counting. Each QA pair provides both OpenQA and CloseQA formats to support fine-grained evaluation. Extensive experiments show that most existing MLLMs, whether general-purpose or egocentric-specialized, struggle to generalize to domains beyond daily life, highlighting the limitations of current models. Furthermore, we conduct several pilot studies, e.g., fine-tuning and reinforcement learning, to explore potential improvements. We hope EgoCross and our accompanying analysis will serve as a foundation for advancing domain-adaptive, robust egocentric video understanding.