[Submitted on 2 Jul 2026 (v1), last revised 18 Nov 2026 (this version, v2)]
Abstract:Localizing targets in $360^circ$ environments requires combining global context with fine-grained evidence despite projection distortion and horizontal wrap-around. Uneven directional coverage compounds these challenges, particularly for seam-adjacent and polar targets. We present SPIRAL-360, a geometry-aware framework with separately trained single-turn and multi-turn models for panoramic localization. Spherically Stratified Replay (SSR) broadens directional coverage through stratified sampling and consistent transformations of panoramas, target annotations, and tool trajectories. TapSphere incorporates spherical distance priors into panoramic visual attention, while Spherical Direction Regularization (SDR) complements coordinate-token prediction with auxiliary direction supervision during training. The single-turn model predicts directly from the panorama, while the multi-turn model inspects tool-generated perspective views. We construct SPIRAL-360 Search, a curated dataset of trajectories interleaving reasoning, projection actions, and visual observations. Both models undergo full-parameter supervised fine-tuning followed by policy optimization. Experiments demonstrate state-of-the-art localization on S360 and competitive cross-dataset transfer, including H$^*$Bench. Systematic ablations support the effectiveness of this geometric learning design.
Submission history
From: Jingtao Xu [view email]
[v1] Thu, 2 Jul 2026 17:47:27 UTC (20,152 KB)
[v2] Wed, 18 Nov 2026 15:03:31 UTC (3,084 KB)
