RePer-360: Releasing Perspective Priors for 360$^\circ$ Depth Estimation via Self-Modulation

• URL: http://arxiv.org/abs/2603.05999v1
• Date: Fri, 06 Mar 2026 07:58:53 GMT
• Title: RePer-360: Releasing Perspective Priors for 360$^\circ$ Depth Estimation via Self-Modulation
• Abstract summary: Recent depth foundation models trained on perspective imagery achieve strong performance, yet generalize poorly to 360$circ$ images.<n>We propose RePer-360, a distortion-aware self-modulation framework for monocular panoramic depth estimation.<n>By shifting the focus from complementary-projection fusion to panoramic domain adaptation under preserved pretrained perspective priors, RePer-360 surpasses standard fine-tuning methods.
• Score: 20.850233169948634
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent depth foundation models trained on perspective imagery achieve strong performance, yet generalize poorly to 360$^\circ$ images due to the substantial geometric discrepancy between perspective and panoramic domains. Moreover, fully fine-tuning these models typically requires large amounts of panoramic data. To address this issue, we propose RePer-360, a distortion-aware self-modulation framework for monocular panoramic depth estimation that adapts depth foundation models while preserving powerful pretrained perspective priors. Specifically, we design a lightweight geometry-aligned guidance module to derive a modulation signal from two complementary projections (i.e., ERP and CP) and use it to guide the model toward the panoramic domain without overwriting its pretrained perspective knowledge. We further introduce a Self-Conditioned AdaLN-Zero mechanism that produces pixel-wise scaling factors to reduce the feature distribution gap between the perspective and panoramic domains. In addition, a cubemap-domain consistency loss further improves training stability and cross-projection alignment. By shifting the focus from complementary-projection fusion to panoramic domain adaptation under preserved pretrained perspective priors, RePer-360 surpasses standard fine-tuning methods while using only 1\% of the training data. Under the same in-domain training setting, it further achieves an approximately 20\% improvement in RMSE. Code will be released upon acceptance.

Related papers

• SphereSOD: Geometry-Structure Coupled Learning for 360 Salient Object Detection [75.77856258186854]
  360 salient object detection aims to accurately segment salient regions across a full field of view.<n>Existing methods mainly focus on compensating projection distortion while overlooking the interaction between panoramic geometry and salient object structure.<n>We propose SphereSOD, an ERP-native framework that couples panoramic geometry with evolving salient structures.
  arXiv  Detail & Related papers  (2026-09-07T14:47:54Z)
• DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images [20.087102336395173]
  We introduce DepthMaster, a unified metric depth estimation framework.<n>We reformulate the problem by decomposing panoramic images into overlapping perspective patches.<n> DepthMaster achieves state-of-the-art zero-shot performance on 13 diverse datasets.
  arXiv  Detail & Related papers  (2026-06-10T17:35:23Z)
• Unified Panoramic Geometry Estimation via Multi-View Foundation Models [22.160964470897003]
  PaGeR (Panoramic Geometry Reconstruction) is a framework to lift powerful 3D foundation models to the panorama domain.<n>Our strategy is to start from a pre-trained transformer for 3D reconstruction and turn it into a unified high-performance model.<n>We extensively test our method in both indoor and outdoor environments and find that it delivers state-of-the-art performance.
  arXiv  Detail & Related papers  (2026-05-25T22:32:42Z)
• VGGT-360: Geometry-Consistent Zero-Shot Panoramic Depth Estimation [18.215186861036607]
  VGGT-360 is a training-free framework for zero-shot, geometry-consistent panoramic depth estimation.<n> VGGT-360 reformulates the task as panoramic reprojection over multi-view reconstructed 3D models.<n>It integrates three plug-and-play modules that form a unified panorama-to-3D-to-depth framework.
  arXiv  Detail & Related papers  (2026-03-19T14:18:17Z)
• World-Shaper: A Unified Framework for 360° Panoramic Editing [57.174341220144605]
  Existing perspective-based image editing methods fail to model the spatial structure of panoramas.<n>We present World-Shaper, a unified geometry-aware framework that bridges panoramic generation and editing within a single editing-centric design.<n>Our method achieves superior geometric consistency, editing fidelity, and text controllability compared to SOTA methods.
  arXiv  Detail & Related papers  (2026-01-30T19:38:54Z)
• Depth Anything in $360^\circ$: Towards Scale Invariance in the Wild [12.6239596554452]
  We present Depth Anything in $360circ$ (DA360), a panoramic-adapted version of Depth Anything V2.<n>Our key innovation involves learning a shift parameter from the ViT backbone, transforming the model's scale- and shift-invariant output into a scale-invariant estimate.
  arXiv  Detail & Related papers  (2025-12-28T07:12:58Z)
• SE360: Semantic Edit in 360$^\circ$ Panoramas via Hierarchical Data Construction [14.137976445056466]
  SE360 is a novel framework for multi-condition guided object editing in 360$circ$ panoramas.<n>At its core is a novel coarse-to-fine autonomous data generation pipeline without manual intervention.<n>Our experiments demonstrate that our method outperforms existing methods in both visual quality and semantic accuracy.
  arXiv  Detail & Related papers  (2025-12-23T00:24:46Z)
• Dual-Projection Fusion for Accurate Upright Panorama Generation in Robotic Vision [9.05196155518077]
  This study presents a dual-stream angle-aware generation network that jointly estimates camera inclination angles and reconstructs upright panoramic images.<n> Experiments on the SUN360 and M3D datasets demonstrate that our method outperforms existing approaches in both inclination estimation and upright panorama generation.
  arXiv  Detail & Related papers  (2025-11-30T14:28:21Z)
• DiT360: High-Fidelity Panoramic Image Generation via Hybrid Training [76.82789568988557]
  DiT360 is a DiT-based framework that performs hybrid training on perspective and panoramic data for panoramic image generation.<n>Our method achieves better boundary consistency and image fidelity across eleven quantitative metrics.
  arXiv  Detail & Related papers  (2025-10-13T17:59:15Z)
• SphereDiff: Tuning-free Omnidirectional Panoramic Image and Video Generation via Spherical Latent Representation [31.305851707485967]
  We introduce SphereDiff, a novel approach for seamless 360-degree panoramic image and video generation using state-of-the-art diffusion models without additional tuning.<n>We extend MultiDiffusion to spherical latent space and propose a spherical latent sampling method to enable direct use of pretrained diffusion models.<n>Our method outperforms existing approaches in generating 360-degree panoramic content while maintaining high fidelity, making it a robust solution for immersive AR/VR applications.
  arXiv  Detail & Related papers  (2025-04-19T19:59:11Z)
• Revisiting 360 Depth Estimation with PanoGabor: A New Fusion Perspective [89.53522682640337]
  We propose an oriented distortion-aware Gabor Fusion framework (PGFuse) to address the above challenges.<n>To address the reintroduced distortions, we design a linear latitude-aware distortion representation method to generate customized, distortion-aware Gabor filters.<n>Considering the orientation sensitivity of the Gabor transform, we introduce a spherical gradient constraint to stabilize this sensitivity.
  arXiv  Detail & Related papers  (2024-08-29T02:58:35Z)
• GEOcc: Geometrically Enhanced 3D Occupancy Network with Implicit-Explicit Depth Fusion and Contextual Self-Supervision [49.839374549646884]
  This paper presents GEOcc, a Geometric-Enhanced Occupancy network tailored for vision-only surround-view perception.<n>Our approach achieves State-Of-The-Art performance on the Occ3D-nuScenes dataset with the least image resolution needed and the most weightless image backbone.
  arXiv  Detail & Related papers  (2024-05-17T07:31:20Z)
• 360 Layout Estimation via Orthogonal Planes Disentanglement and Multi-view Geometric Consistency Perception [56.84921040837699]
  Existing panoramic layout estimation solutions tend to recover room boundaries from a vertically compressed sequence, yielding imprecise results. We propose an orthogonal plane disentanglement network (termed DOPNet) to distinguish ambiguous semantics. We also present an unsupervised adaptation technique tailored for horizon-depth and ratio representations. Our solution outperforms other SoTA models on both monocular layout estimation and multi-view layout estimation tasks.
  arXiv  Detail & Related papers  (2023-12-26T12:16:03Z)

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.