OmniEgo-R$^2$: A Routed Reasoning Framework for the 1st Cross-Domain EgoCross Challenge at CVPR 2026

• URL: http://arxiv.org/abs/2605.24481v2
• Date: Wed, 27 May 2026 06:08:16 GMT
• Title: OmniEgo-R$^2$: A Routed Reasoning Framework for the 1st Cross-Domain EgoCross Challenge at CVPR 2026
• Abstract summary: We formulate EgoCross as a robust cross-domain embodied video reasoning problem.<n>We identify three key challenges: temporal boundary ambiguity, cross-domain semantic mismatch, and decision instability under close options.<n>Our submissions obtain 66.35% overall accuracy in the Source-Limited track and 66.77% in the Open-Source track, ranking second in both leaderboards.
• Score: 73.31700707400647
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The 1st Cross-Domain EgoCross Challenge at EgoVis, CVPR 2026 evaluates whether multimodal large language models can reason over egocentric videos across surgery, industry, extreme sports, and animal perspective. We achieved second place in both the Source-Limited and Open-Source tracks. In this report, we formulate EgoCross as a robust cross-domain embodied video reasoning problem rather than a simple multiple-choice visual question answering task. We identify three key challenges: (C1) temporal boundary ambiguity, where critical state transitions are sparsely sampled and often occur between frames; (C2) cross-domain semantic granularity mismatch, where the same capability requires different domain-specific visual grammar; and (C3) decision instability under close options, where long multimodal reasoning can select unsupported distractors or produce malformed outputs. To address them, we propose OmniEgo-R$^2$ (Omnidomain Egocentric Routed Reasoning), a unified routed reasoning pipeline consisting of temporal-evidence normalization, domain-agnostic capability routing, structured perception--dynamics--decision reasoning, boundary-aware option verification, and defensive answer calibration. OmniEgo-R$^2$ uses the Qwen3-VL-4B-SFT checkpoints on each EgoCross domain as the visual-language backbone, and wraps them with lightweight test-time reasoning and parsing programs. Our final submissions obtain 66.35% overall accuracy in the Source-Limited track and 66.77% in the Open-Source track, ranking second in both leaderboards. The codes are available on https://github.com/Lee-zixu/OmniEgo-R2

Related papers

• The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering [196.78334982965507]
  EgoCross is a cross-domain egocentric video question answering benchmark.<n>It is designed to evaluate whether multimodal large language models can generalize beyond common daily-life scenarios.<n>The first EgoCross Challenge was hosted at the Third EgoVis Workshop at CVPR 2026.
  arXiv  Detail & Related papers  (2026-08-05T08:51:20Z)
• PixDLM: A Dual-Path Multimodal Language Model for UAV Reasoning Segmentation [71.53011159355401]
  Reasoning segmentation has expanded from ground-level scenes to remote-sensing imagery, yet UAV data poses distinct challenges.<n>We define the UAV Reasoning task and organize its semantic requirements into three dimensions: Spatial, Attribute, and Scene-level reasoning.<n>We construct DRSeg, a large-scale benchmark for UAV reasoning segmentation, containing 10k high-resolution aerial images paired with Chain-of-Thought QA supervision.
  arXiv  Detail & Related papers  (2026-04-17T03:48:56Z)
• DriveXQA: Cross-modal Visual Question Answering for Adverse Driving Scene Understanding [40.93547456625219]
  We propose the DriveXQA, a multimodal dataset for autonomous driving VQA.<n>In addition to four visual modalities, five sensor failure cases, and five weather conditions, it includes $102,505$ QA pairs categorized into three types: global scene level, allocentric level, and ego-vehicle centric level.<n>We design MVX-LLM, a token-efficient architecture with a Dual Cross-Attention (DCA) projector that fuses the modalities to alleviate information redundancy.
  arXiv  Detail & Related papers  (2026-03-11T23:49:11Z)
• EgoThinker: Unveiling Egocentric Reasoning with Spatio-Temporal CoT [56.24624833924252]
  EgoThinker is a framework that endows MLs with robust egocentric reasoning capabilities through-temporal chain-of-thought supervision and a two-stage learning curriculum.<n>EgoThinker outperforms existing methods across multiple egocentric benchmarks, while achieving substantial improvements in fine-grained-temporal localization tasks.
  arXiv  Detail & Related papers  (2025-10-27T17:38:17Z)
• ERGO: Efficient High-Resolution Visual Understanding for Vision-Language Models [11.263321053154364]
  ERGO is a reasoning-driven perception-leveraging multimodal context to determine where to focus.<n>We develop simple yet effective reward components in a reinforcement learning framework for coarse-to-fine perception.<n>Our approach delivers higher accuracy than the original model and competitive methods, with greater efficiency.
  arXiv  Detail & Related papers  (2025-09-26T07:15:19Z)
• EgoCross: Benchmarking Multimodal Large Language Models for Cross-Domain Egocentric Video Question Answering [59.94048858464922]
  EgoCross is a benchmark to evaluate the cross-domain generalization of MLLMs in EgocentricQA.<n>EgoCross covers four diverse and challenging domains, including surgery, industry, extreme sports, and animal perspective.<n>It comprises approximately 1,000 QA pairs across 798 video clips, spanning four key QA tasks: prediction, recognition, localization, and counting.
  arXiv  Detail & Related papers  (2025-08-14T15:11:20Z)
• Traceable Evidence Enhanced Visual Grounded Reasoning: Evaluation and Methodology [87.65242416688146]
  TreeBench is a diagnostic benchmark for visual grounded reasoning.<n>TreeVGR is a training paradigm to supervise localization and reasoning jointly with reinforcement learning.
  arXiv  Detail & Related papers  (2025-07-10T17:59:58Z)
• One for All: Multi-Domain Joint Training for Point Cloud Based 3D Object Detection [71.78795573911512]
  We propose textbfOneDet3D, a universal one-for-all model that addresses 3D detection across different domains. We propose the domain-aware in scatter and context, guided by a routing mechanism, to address the data interference issue. The fully sparse structure and anchor-free head further accommodate point clouds with significant scale disparities.
  arXiv  Detail & Related papers  (2024-11-03T14:21:56Z)
• X-InstructBLIP: A Framework for aligning X-Modal instruction-aware representations to LLMs and Emergent Cross-modal Reasoning [109.9413329636322]
  This paper introduces an efficient framework that integrates multiple modalities (images, 3D, audio and video) to a frozen Large Language Models (LLMs) Our approach explores two distinct projection mechanisms: Q-Formers and Linear Projections (LPs)
  arXiv  Detail & Related papers  (2023-11-30T18:43:51Z)
• DualCross: Cross-Modality Cross-Domain Adaptation for Monocular BEV Perception [30.113617846516398]
  DualCross is a cross-modality cross-domain adaptation framework to facilitate the learning of a more robust BEV perception model. This work results in the first open analysis of cross-domain cross-sensor perception and adaptation for monocular 3D tasks in the wild.
  arXiv  Detail & Related papers  (2023-05-05T17:58:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.