EAGLE-360: Embodied Active Global-to-Local Exploration in 360$^\circ$

• URL: http://arxiv.org/abs/2607.02479v1
• Date: Thu, 02 Jul 2026 17:47:27 GMT
• Title: EAGLE-360: Embodied Active Global-to-Local Exploration in 360$^\circ$
• Abstract summary: We propose a novel Embodied Active Global-to-Local Exploration framework for 360$circ$ visual search.<n>We construct a large-scale dataset, comprising 14,000+ 4K panoramas and 70,000+ rounds of high-quality VQA dialogues.<n>Experiments demonstrate that the framework achieves nearly an 8-fold increase in accuracy over the base model.
• Score: 25.40870252692496
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: While Multimodal Large Language Models (MLLMs) have demonstrated exceptional capabilities in standard visual understanding, adapting them for active visual search in 360$^\circ$ panoramic environments exposes fundamental limitations. Specifically, standard MLLMs struggle to effectively model inherent panoramic properties, such as severe polar distortion and continuous cylindrical topologies, which significantly degrades target detection accuracy. Consequently, existing panoramic search methods attempt to compensate by relying heavily on fragmented local viewpoints. Burdened by rigid initialization and a lack of global panoramic priors, these approaches suffer from myopic, inefficient exploration and struggle with robust error recovery when targets are out of view. To overcome these challenges, we propose EAGLE-360, a novel Embodied Active Global-to-Local Exploration framework. Rather than performing exhaustive local searches, EAGLE-360 leverages global priors to establish an initial holistic perspective, iteratively reasoning and progressively narrowing the search space. Architecturally, we adapt RoPE Rolling, a coordinate-shifting positional encoding mechanism, to seamlessly model the continuous topologies of panoramas. To facilitate this paradigm, we construct the large-scale EAGLE-360 dataset, comprising 14,000+ 4K panoramas and 70,000+ rounds of high-quality VQA dialogues. By employing a training pipeline that integrates Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO), we effectively elicit complex spatial reasoning and tool-calling capabilities. Extensive experiments demonstrate that EAGLE-360 establishes a new state-of-the-art for 360$^\circ$ visual search, achieving nearly an 8-fold increase in accuracy over the base model while significantly enhancing exploration efficiency.

Related papers

• PanoWorld: Real-World Panoramic Generation [68.3569360188378]
  We propose PanoWorld, which simplifies camera trajectories into translations via fixed headings for both current-action modeling and long-range memory.<n>Extensive experiments on World360 demonstrate the effectiveness of PanoWorld, outperforming alternative methods by a large margin.
  arXiv  Detail & Related papers  (2026-07-10T17:59:32Z)
• Seek to Segment: Active Perception for Panoramic Referring Segmentation [76.39419654494505]
  Existing referring segmentation models passively process static images captured from fixed perspectives.<n>We propose PanoSeeker, a memory-augmented agent for efficient Active Panoramic Referring.<n>Experiments on our newly established APRS benchmark demonstrate that PanoSeeker achieves superior search efficiency and segmentation accuracy.
  arXiv  Detail & Related papers  (2026-07-02T17:56:49Z)
• OmniCoT: A Benchmark for Global and Multi-Step Panoramic Reasoning [94.47284498900895]
  We introduce OmniCoT, a panoramic spatial reasoning suite designed to enable MLLMs to use global evidence and perform multi-step inference across viewpoints.<n>For training, OmniCoT-T is purpose-built with structured stepwise Chain-of-Thought annotations that explicitly link intermediate reasoning steps to panoramic evidence.
  arXiv  Detail & Related papers  (2026-06-29T14:38:20Z)
• Panoramic Affordance Prediction [94.50813972018504]
  Affordance prediction serves as a critical bridge between perception and action in embodied AI.<n>We present the first exploration into Panoramic Affordance Prediction, utilizing 360-degree imagery.<n>We propose PAP, a training-free, coarse-to-fine pipeline inspired by the human foveal visual system.
  arXiv  Detail & Related papers  (2026-03-16T17:21:49Z)
• PanoAffordanceNet: Towards Holistic Affordance Grounding in 360° Indoor Environments [18.94879808515964]
  We introduce a novel task: Holistic Affordance Grounding in 360 Indoor Environments.<n>This task faces unique challenges, including severe geometric distortions from Equirectangular Projection (ERP), semantic dispersion, and cross-scale alignment difficulties.<n>We propose PanoAffordanceNet, an end-to-end framework featuring a Distortion-Aware Spectral Modulator (DASM) for latitude-dependent calibration and an Omni-Spherical Densification Head (OSDH) to restore topological continuity from sparse activations.
  arXiv  Detail & Related papers  (2026-03-10T14:58:26Z)
• TagaVLM: Topology-Aware Global Action Reasoning for Vision-Language Navigation [70.23578202012048]
  Vision-Language Navigation (VLN) presents a unique challenge for Large Vision-Language Models (VLMs) due to their inherent architectural mismatch.<n>We propose TagaVLM (Topology-Aware Global Action reasoning), an end-to-end framework that explicitly injects topological structures into the VLM backbone.<n>To enhance topological node information, an Interleaved Navigation Prompt strengthens node-level visual-text alignment.<n>With the embedded topological graph, the model is capable of global action reasoning, allowing for robust path correction.
  arXiv  Detail & Related papers  (2026-03-03T13:28:07Z)
• ZoomEarth: Active Perception for Ultra-High-Resolution Geospatial Vision-Language Tasks [49.99788276124186]
  Existing dynamic resolution and token pruning methods are constrained by a passive perception paradigm.<n>We present LRS-GRO, a large-scale benchmark dataset tailored for active perception in UHR RS processing.<n>We propose ZoomEarth, an adaptive cropping-zooming framework with a novel Region-Guided reward that provides fine-grained guidance.
  arXiv  Detail & Related papers  (2025-11-15T15:47:46Z)
• Omnidirectional Spatial Modeling from Correlated Panoramas [4.75637997496421]
  Existing omnidirectional methods achieve scene understanding within a single frame while neglecting cross-frame correlated panoramas.<n>We introduce textbfCFpano, the textbffirst benchmark dataset dedicated to cross-frame correlated panoramas visual question answering.<n>We present methodname, a multi-modal large language model (MLLM) fine-tuned with Group Relative Policy Optimization ( GRPO) and a set of tailored reward functions for robust and consistent reasoning with cross-frame correlated panoramas.
  arXiv  Detail & Related papers  (2025-09-02T10:14:55Z)
• ACT360: An Efficient 360-Degree Action Detection and Summarization Framework for Mission-Critical Training and Debriefing [3.781421673607643]
  We introduce ACT360, a system that leverages 360-degree videos and machine learning for automated action detection and structured debriefing.<n>ACT360 integrates 360YOWO, an enhanced You Only Watch Once (YOWO) model with spatial attention and equirectangular-aware convolution (EAC) to mitigate panoramic video distortions.<n>We validate our approach on a publicly available dataset of 55 labeled 360-degree videos covering seven key operational actions.
  arXiv  Detail & Related papers  (2025-03-17T06:12:36Z)
• Centralized Feature Pyramid for Object Detection [53.501796194901964]
  Visual feature pyramid has shown its superiority in both effectiveness and efficiency in a wide range of applications. In this paper, we propose a OLO Feature Pyramid for object detection, which is based on a globally explicit centralized feature regulation.
  arXiv  Detail & Related papers  (2022-10-05T08:32:54Z)

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.