PanDA: Towards Panoramic Depth Anything with Unlabeled Panoramas and Mobius Spatial Augmentation

• URL: http://arxiv.org/abs/2406.13378v2
• Date: Sat, 15 Mar 2025 09:07:43 GMT
• Title: PanDA: Towards Panoramic Depth Anything with Unlabeled Panoramas and Mobius Spatial Augmentation
• Authors: Zidong Cao, Jinjing Zhu, Weiming Zhang, Hao Ai, Haotian Bai, Hengshuang Zhao, Lin Wang,
• Abstract summary: We propose a semi-supervised learning framework to learn a panoramic depth foundation model, dubbed PanDA.<n>PanDA learns a teacher model by fine-tuning DAM through joint training on synthetic indoor and outdoor panoramic datasets.<n>Experiments demonstrate that PanDA exhibits remarkable zero-shot capability across diverse scenes.
• Score: 35.206582458026574
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recently, Depth Anything Models (DAMs) - a type of depth foundation models - have demonstrated impressive zero-shot capabilities across diverse perspective images. Despite its success, it remains an open question regarding DAMs' performance on panorama images that enjoy a large field-of-view (180x360) but suffer from spherical distortions. To address this gap, we conduct an empirical analysis to evaluate the performance of DAMs on panoramic images and identify their limitations. For this, we undertake comprehensive experiments to assess the performance of DAMs from three key factors: panoramic representations, 360 camera positions for capturing scenarios, and spherical spatial transformations. This way, we reveal some key findings, e.g., DAMs are sensitive to spatial transformations. We then propose a semi-supervised learning (SSL) framework to learn a panoramic DAM, dubbed PanDA. Under the umbrella of SSL, PanDA first learns a teacher model by fine-tuning DAM through joint training on synthetic indoor and outdoor panoramic datasets. Then, a student model is trained using large-scale unlabeled data, leveraging pseudo-labels generated by the teacher model. To enhance PanDA's generalization capability, M"obius transformation-based spatial augmentation (MTSA) is proposed to impose consistency regularization between the predicted depth maps from the original and spatially transformed ones. This subtly improves the student model's robustness to various spatial transformations, even under severe distortions. Extensive experiments demonstrate that PanDA exhibits remarkable zero-shot capability across diverse scenes, and outperforms the data-specific panoramic depth estimation methods on two popular real-world benchmarks.

Related papers

• DiffPano: Scalable and Consistent Text to Panorama Generation with Spherical Epipolar-Aware Diffusion [60.45000652592418]
  We propose a novel text-driven panoramic generation framework, DiffPano, to achieve scalable, consistent, and diverse panoramic scene generation. We show that DiffPano can generate consistent, diverse panoramic images with given unseen text descriptions and camera poses.
  arXiv  Detail & Related papers  (2024-10-31T17:57:02Z)
• Depth Anywhere: Enhancing 360 Monocular Depth Estimation via Perspective Distillation and Unlabeled Data Augmentation [6.832852988957967]
  We propose a new depth estimation framework that utilizes unlabeled 360-degree data effectively. Our approach uses state-of-the-art perspective depth estimation models as teacher models to generate pseudo labels. We tested our approach on benchmark datasets such as Matterport3D and Stanford2D3D, showing significant improvements in depth estimation accuracy.
  arXiv  Detail & Related papers  (2024-06-18T17:59:31Z)
• Sp2360: Sparse-view 360 Scene Reconstruction using Cascaded 2D Diffusion Priors [51.36238367193988]
  We tackle sparse-view reconstruction of a 360 3D scene using priors from latent diffusion models (LDM) We present SparseSplat360, a method that employs a cascade of in-painting and artifact removal models to fill in missing details and clean novel views. Our method generates entire 360 scenes from as few as 9 input views, with a high degree of foreground and background detail.
  arXiv  Detail & Related papers  (2024-05-26T11:01:39Z)
• 360SFUDA++: Towards Source-free UDA for Panoramic Segmentation by Learning Reliable Category Prototypes [15.367186190755003]
  We address the challenging source-free unsupervised domain adaptation (SFUDA) for pinhole-to-panoramic semantic segmentation. 360SFUDA++ effectively extracts knowledge from the source pinhole model with only unlabeled panoramic images.
  arXiv  Detail & Related papers  (2024-04-25T10:52:08Z)
• Leveraging Large-Scale Pretrained Vision Foundation Models for Label-Efficient 3D Point Cloud Segmentation [67.07112533415116]
  We present a novel framework that adapts various foundational models for the 3D point cloud segmentation task. Our approach involves making initial predictions of 2D semantic masks using different large vision models. To generate robust 3D semantic pseudo labels, we introduce a semantic label fusion strategy that effectively combines all the results via voting.
  arXiv  Detail & Related papers  (2023-11-03T15:41:15Z)
• Distortion-aware Transformer in 360{\deg} Salient Object Detection [44.74647420381127]
  We propose a Transformer-based model called DATFormer to address the distortion problem. To exploit the unique characteristics of 360deg data, we present a learnable relation matrix. Our model outperforms existing 2D SOD (salient object detection) and 360 SOD methods.
  arXiv  Detail & Related papers  (2023-08-07T07:28:24Z)
• PanoDiffusion: 360-degree Panorama Outpainting via Diffusion [29.10215003547589]
  We present our 360-degree indoor RGB-D panorama outpainting model using latent diffusion models (LDM), called PanoDiffusion. We introduce a new bi-modal latent diffusion structure that utilizes both RGB and depth panoramic data during training, which works surprisingly well to outpaint depth-free RGB images during inference. Results show that our PanoDiffusion not only significantly outperforms state-of-the-art methods on RGB-D panorama outpainting by producing diverse well-structured results for different types of masks, but can also synthesize high-quality depth panoramas to provide realistic 3D indoor models.
  arXiv  Detail & Related papers  (2023-07-06T17:57:02Z)
• SGAT4PASS: Spherical Geometry-Aware Transformer for PAnoramic Semantic Segmentation [53.5256153325136]
  PAnoramic Semantic (PASS) gives complete scene perception based on an ultra-wide angle of view. Usually, prevalent PASS methods with 2D panoramic image input focus on solving image distortions but lack consideration of the 3D properties of original $360circ$ data. We propose Spherical Geometry-Aware Transformer for PAnoramic Semantic (SGAT4PASS) to be more robust to 3D disturbance.
  arXiv  Detail & Related papers  (2023-06-06T04:49:51Z)
• PanoGRF: Generalizable Spherical Radiance Fields for Wide-baseline Panoramas [54.4948540627471]
  We propose PanoGRF, Generalizable Spherical Radiance Fields for Wide-baseline Panoramas. Unlike generalizable radiance fields trained on perspective images, PanoGRF avoids the information loss from panorama-to-perspective conversion. Results on multiple panoramic datasets demonstrate that PanoGRF significantly outperforms state-of-the-art generalizable view synthesis methods.
  arXiv  Detail & Related papers  (2023-06-02T13:35:07Z)
• NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with 360{\deg} Views [77.93662205673297]
  In this work, we study the challenging task of lifting a single image to a 3D object. We demonstrate the ability to generate a plausible 3D object with 360deg views that correspond well with a given reference image. We propose a novel framework, dubbed NeuralLift-360, that utilizes a depth-aware radiance representation.
  arXiv  Detail & Related papers  (2022-11-29T17:59:06Z)
• View-aware Salient Object Detection for 360{\deg} Omnidirectional Image [33.43250302656753]
  We construct a large scale 360deg ISOD dataset with object-level pixel-wise annotation on equirectangular projection (ERP) Inspired by humans' observing process, we propose a view-aware salient object detection method based on a Sample Adaptive View Transformer (SAVT) module.
  arXiv  Detail & Related papers  (2022-09-27T07:44:08Z)
• PanoFormer: Panorama Transformer for Indoor 360{\deg} Depth Estimation [35.698249161263966]
  Existing panoramic depth estimation methods based on convolutional neural networks (CNNs) focus on removing panoramic distortions. This paper proposes the panorama transformer (named PanoFormer) to estimate the depth in panorama images. In particular, we divide patches on the spherical tangent domain into tokens to reduce the negative effect of panoramic distortions.
  arXiv  Detail & Related papers  (2022-03-17T12:19:43Z)
• Bending Reality: Distortion-aware Transformers for Adapting to Panoramic Semantic Segmentation [26.09267582056609]
  Large quantities of expensive, pixel-wise annotations are crucial for success of robust panoramic segmentation models. Distortions and the distinct image-feature distribution in 360-degree panoramas impede the transfer from the annotation-rich pinhole domain. We learn object deformations and panoramic image distortions in Deformable Patch Embedding (DPE) and Deformable Deformable (DMLP) components. Finally, we tie together shared semantics in pinhole- and panoramic feature embeddings by generating multi-scale prototype features.
  arXiv  Detail & Related papers  (2022-03-02T23:00:32Z)
• PANet: Perspective-Aware Network with Dynamic Receptive Fields and Self-Distilling Supervision for Crowd Counting [63.84828478688975]
  We propose a novel perspective-aware approach called PANet to address the perspective problem. Based on the observation that the size of the objects varies greatly in one image due to the perspective effect, we propose the dynamic receptive fields (DRF) framework. The framework is able to adjust the receptive field by the dilated convolution parameters according to the input image, which helps the model to extract more discriminative features for each local region.
  arXiv  Detail & Related papers  (2021-10-31T04:43:05Z)
• SHD360: A Benchmark Dataset for Salient Human Detection in 360{\deg} Videos [26.263614207849276]
  We propose SHD360, the first 360deg video SHD dataset collecting various real-life daily scenes. SHD360 contains 16,238 salient human instances with manually annotated pixel-wise ground truth. Our proposed dataset and benchmark could serve as a good starting point for advancing human-centric researches towards 360deg panoramic data.
  arXiv  Detail & Related papers  (2021-05-24T23:51:29Z)
• Panoramic Panoptic Segmentation: Towards Complete Surrounding Understanding via Unsupervised Contrastive Learning [97.37544023666833]
  We introduce panoramic panoptic segmentation as the most holistic scene understanding. A complete surrounding understanding provides a maximum of information to the agent. We propose a framework which allows model training on standard pinhole images and transfers the learned features to a different domain.
  arXiv  Detail & Related papers  (2021-03-01T09:37:27Z)
• A Fixation-based 360{\deg} Benchmark Dataset for Salient Object Detection [21.314578493964333]
  Fixation prediction (FP) in panoramic contents has been widely investigated along with the booming trend of virtual reality (VR) applications. salient object detection (SOD) has been seldom explored in 360deg images due to the lack of datasets representative of real scenes.
  arXiv  Detail & Related papers  (2020-01-22T11:16:39Z)
• Visual Question Answering on 360{\deg} Images [96.00046925811515]
  VQA 360 is a novel task of visual question answering on 360 images. We collect the first VQA 360 dataset, containing around 17,000 real-world image-question-answer triplets for a variety of question types.
  arXiv  Detail & Related papers  (2020-01-10T08:18:21Z)

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.