Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
38 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
41 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Customizable Perturbation Synthesis for Robust SLAM Benchmarking (2402.08125v1)

Published 12 Feb 2024 in cs.RO, cs.AI, cs.CV, and cs.MM
Customizable Perturbation Synthesis for Robust SLAM Benchmarking

Abstract: Robustness is a crucial factor for the successful deployment of robots in unstructured environments, particularly in the domain of Simultaneous Localization and Mapping (SLAM). Simulation-based benchmarks have emerged as a highly scalable approach for robustness evaluation compared to real-world data collection. However, crafting a challenging and controllable noisy world with diverse perturbations remains relatively under-explored. To this end, we propose a novel, customizable pipeline for noisy data synthesis, aimed at assessing the resilience of multi-modal SLAM models against various perturbations. This pipeline incorporates customizable hardware setups, software components, and perturbed environments. In particular, we introduce comprehensive perturbation taxonomy along with a perturbation composition toolbox, allowing the transformation of clean simulations into challenging noisy environments. Utilizing the pipeline, we instantiate the Robust-SLAM benchmark, which includes diverse perturbation types, to evaluate the risk tolerance of existing advanced multi-modal SLAM models. Our extensive analysis uncovers the susceptibilities of existing SLAM models to real-world disturbance, despite their demonstrated accuracy in standard benchmarks. Our perturbation synthesis toolbox, SLAM robustness evaluation pipeline, and Robust-SLAM benchmark will be made publicly available at https://github.com/Xiaohao-Xu/SLAM-under-Perturbation/.

PDF HTML Abstract

An Examination of the Placeholder Document Structure in LaTeX

This brief document embodies a minimal LaTeX structure, often utilized as a foundational template for academic papers. It exhibits several core components aligned with established norms in academic publishing. The provided content constitutes the skeleton of a LaTeX document, specifically highlighting configuration aspects relevant to document preparation, rather than substantive research findings.

Core Components and Configuration

The document begins with a class declaration: \documentclass{article}. This choice indicates that the document will adopt the formatting conventions standard to articles, a common option in scholarly publications due to its simplicity and suitability for text-heavy manuscripts. The primary purpose of this declaration lies in delivering consistent typographical standards across the document, allowing authors to maintain uniformity in text presentation.

Following the class declaration, the document advances into the content environment with \begin{document} and concludes with \end{document}. These commands demarcate the boundaries of the document's main content, encapsulating all sections intended for print or viewing within a LaTeX processing system.

A notable feature within the document is the bibliographic-related content. The command \nocite{*} is employed to include all references in the associated bibliography, irrespective of their citation status within the document. This can be advantageous during the drafting phases of an academic paper when researchers may wish to compile a comprehensive list of literature for later refinement.

The line \bibliographystyle{plainnat} specifies the citation style to be used, indicating a preference for the plainnat style associated with the natbib package. This style plainnat offers flexibility for citing in author-year format, facilitating a nuanced approach to bibliographic entry configuration.

Implications for Academic Publishing

Though this document serves no explicit research purpose on its own, its structural and procedural elements provide value to researchers preparing manuscripts. Proper configuration of a LaTeX document ensures seamless compilation, professional appearance, and adherence to the rigorous standards often required by publishers.

By utilizing this fundamental setup, researchers can concentrate on writing and research contributions while relying on LaTeX to optimize formatting and presentation. Moreover, it opens opportunities for further customization, enabling users to integrate additional packages that can accommodate specific requirements such as advanced mathematical notation, algorithm layout, or graphic inclusion.

Future Directions in Document Preparation

The simplicity of this template provides a starting point for exploring more intricate document structures. Future enhancements might consider the integration of collaborative features or automated formatting checks that align with evolving publication standards. Additionally, developments in AI-assisted writing tools could further simplify the document preparation process by suggesting optimizations or corrections in real-time.

In conclusion, while the provided LaTeX document does not deliver empirical research, it encapsulates the functionality and foundational consistency necessary for the preparation of academic texts. This serves as a testament to LaTeX’s enduring role in the academic domain and offers a preview into evolving document processing innovations that cater to the scholarly community.

File Document Download Save Streamline Icon: https://streamlinehq.com PDF File Document Download Save Streamline Icon: https://streamlinehq.com Markdown Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (81)
  1. Task-aware risk estimation of perception failures for autonomous vehicles. arXiv preprint arXiv:2305.01870, 2023.
  2. Strategies for reducing speckle noise in digital holography. Light: Science and Applications, 7(1):48, 2018.
  3. The málaga urban dataset: High-rate stereo and lidar in a realistic urban scenario. Int. J. Robot. Research, 33(2):207–214, 2014.
  4. Robust slam systems: Are we there yet? In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5320–5327. IEEE, 2021.
  5. The EuRoC micro aerial vehicle datasets. The International Journal of Robotics Research, 2016.
  6. Past, present, and future of simultaneous localization and mapping: Toward the robust-perception age. IEEE Transactions on Robotics, 32(6):1309–1332, 2016.
  7. Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam. IEEE Transactions on Robotics, 37(6):1874–1890, 2021.
  8. University of michigan north campus long-term vision and lidar dataset. The International Journal of Robotics Research, 35(9):1023–1035, 2016.
  9. Modeling camera effects to improve visual learning from synthetic data. In Proceedings of the European Conference on Computer Vision (ECCV) Workshops, September 2018.
  10. Robustnav: Towards benchmarking robustness in embodied navigation. In IEEE/CVF International Conference on Computer Vision, pages 15691–15700, 2021.
  11. Deep learning for visual localization and mapping: A survey. IEEE Transactions on Neural Networks and Learning Systems, pages 1–21, 2023.
  12. Kaist multi-spectral day/night data set for autonomous and assisted driving. IEEE Transactions on Intelligent Transportation Systems, 19(3):934–948, 2018.
  13. Pearson correlation coefficient. Noise reduction in speech processing, pages 1–4, 2009.
  14. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5828–5839, 2017.
  15. Procthor: Large-scale embodied ai using procedural generation. Advances in Neural Information Processing Systems, 35:5982–5994, 2022.
  16. Are we ready for autonomous drone racing? the UZH-FPV drone racing dataset. In IEEE Int. Conf. Robot. Autom. (ICRA), 2019.
  17. Blenderproc: Reducing the reality gap with photorealistic rendering. In International Conference on Robotics: Sciene and Systems, RSS 2020, 2020.
  18. Benchmarking robustness of 3d object detection to common corruption. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1022–1032, 2023.
  19. Carla: An open urban driving simulator. In Conference on robot learning, pages 1–16. PMLR, 2017.
  20. Present and future of slam in extreme environments: The darpa subt challenge. IEEE Transactions on Robotics, pages 1–20, 2023.
  21. Are we ready for autonomous driving? The KITTI Vision Benchmark Suite. In Conference on Computer Vision and Pattern Recognition (CVPR), 2012.
  22. Samuel W Hasinoff. Photon, poisson noise. Computer Vision, A Reference Guide, 4(16):1, 2014.
  23. The hilti slam challenge dataset. IEEE Robotics and Automation Letters, 7(3):7518–7525, 2022.
  24. Benchmarking neural network robustness to common corruptions and perturbations. Proceedings of the International Conference on Learning Representations, 2019.
  25. Driving in the matrix: Can virtual worlds replace human-generated annotations for real world tasks? In 2017 IEEE International Conference on Robotics and Automation (ICRA), pages 746–753, 2017.
  26. Benchmarking the robustness of semantic segmentation models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8828–8838, 2020a.
  27. Increasing the robustness of semantic segmentation models with painting-by-numbers. In European Conference on Computer Vision, pages 369–387. Springer, 2020b.
  28. Champion-level drone racing using deep reinforcement learning. Nature, 620(7976):982–987, 2023.
  29. A survey of state-of-the-art on visual slam. Expert Systems with Applications, 205:117734, 2022.
  30. Splatam: Splat, track map 3d gaussians for dense rgb-d slam. arXiv, 2023.
  31. 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42(4), 2023.
  32. Mimosa: A multi-modal slam framework for resilient autonomy against sensor degradation. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 7153–7159, 2022.
  33. Tae Kyun Kim. T test as a parametric statistic. Korean journal of anesthesiology, 68(6):540–546, 2015.
  34. Robo3d: Towards robust and reliable 3d perception against corruptions. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 19994–20006, 2023.
  35. Robust referring video object segmentation with cyclic structural consensus. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 22236–22245, 2023.
  36. Homogeneous multi-modal feature fusion and interaction for 3d object detection. In European Conference on Computer Vision, pages 691–707, 2022.
  37. A comprehensive survey of visual slam algorithms. Robotics, 11(1):24, 2022.
  38. 1 year, 1000 km: The oxford robotcar dataset. The International Journal of Robotics Research, 36(1):3–15, 2017.
  39. Benchmarking robustness in object detection: Autonomous driving when winter is coming. arXiv preprint arXiv:1907.07484, 2019.
  40. Universal adversarial perturbations. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1765–1773, 2017.
  41. Extreme low-light environment-driven image denoising over permanently shadowed lunar regions with a physical noise model. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6317–6327, 2021.
  42. Orb-slam: a versatile and accurate monocular slam system. IEEE transactions on robotics, 31(5):1147–1163, 2015.
  43. Orb-slam2: An open-source slam system for monocular, stereo, and rgb-d cameras. IEEE Transactions on Robotics, 33(5):1255–1262, 2017.
  44. Covins-g: A generic back-end for collaborative visual-inertial slam. In 2023 IEEE International Conference on Robotics and Automation (ICRA), pages 2076–2082, 2023.
  45. Penncosyvio: A challenging visual inertial odometry benchmark. In 2017 IEEE International Conference on Robotics and Automation (ICRA), pages 3847–3854. IEEE, 2017.
  46. A survey on active simultaneous localization and mapping: State of the art and new frontiers. IEEE Transactions on Robotics, 39(3):1686–1705, 2023.
  47. Infinite photorealistic worlds using procedural generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12630–12641, 2023.
  48. Zero-shot text-to-image generation. In International Conference on Machine Learning, pages 8821–8831. PMLR, 2021.
  49. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022.
  50. Kimera: From slam to spatial perception with 3d dynamic scene graphs. The International Journal of Robotics Research, 40(12-14):1510–1546, 2021.
  51. Nerf-slam: Real-time dense monocular slam with neural radiance fields. arXiv preprint arXiv:2210.13641, 2022.
  52. Visual-inertial navigation algorithm development using photorealistic camera simulation in the loop. In 2018 IEEE International Conference on Robotics and Automation (ICRA), pages 2566–2573, 2018.
  53. The tum vi benchmark for evaluating visual-inertial odometry. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 1680–1687, 2018.
  54. Dave Shreiner et al. OpenGL programming guide: the official guide to learning OpenGL, versions 3.0 and 3.1. Pearson Education, 2009.
  55. Alvy Ray Smith. Image compositing fundamentals. Microsoft Corporation, 5, 1995.
  56. Charles Spearman. The proof and measurement of association between two things. 1961.
  57. The replica dataset: A digital replica of indoor spaces. arXiv preprint arXiv:1906.05797, 2019.
  58. A benchmark for the evaluation of rgb-d slam systems. In 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 573–580, 2012.
  59. imap: Implicit mapping and positioning in real-time. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 6229–6238, 2021.
  60. Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras. Advances in neural information processing systems, 34:16558–16569, 2021.
  61. Sebastian Thrun. Probabilistic robotics. Commun. ACM, 45(3):52–57, 2002.
  62. Kimera-multi: Robust, distributed, dense metric-semantic slam for multi-robot systems. IEEE Transactions on Robotics, 38(4):2022–2038, 2022.
  63. Resilient and distributed multi-robot visual slam: Datasets, experiments, and lessons learned. arXiv preprint arXiv:2304.04362, 2023.
  64. Mujoco: A physics engine for model-based control. In 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 5026–5033, 2012.
  65. The raincouver scene parsing benchmark for self-driving in adverse weather and at night. IEEE Robotics and Automation Letters, 2(4):2188–2193, 2017.
  66. Simulating photo-realistic snow and fog on existing images for enhanced cnn training and evaluation. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC), pages 41–46. IEEE, 2019.
  67. Co-slam: Joint coordinate and sparse parametric encodings for neural real-time slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13293–13302, 2023.
  68. Tartanair: A dataset to push the limits of visual slam. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 4909–4916. IEEE, 2020.
  69. Tartanvo: A generalizable learning-based vo. In Conference on Robot Learning, pages 1761–1772. PMLR, 2021.
  70. D2 slam: Decentralized and distributed collaborative visual-inertial slam system for aerial swarm. arXiv preprint arXiv:2211.01538, 2022a.
  71. Towards robust video object segmentation with adaptive object calibration. In Proceedings of the 30th ACM International Conference on Multimedia, pages 2709–2718, 2022b.
  72. Benchmarking augmentation methods for learning robust navigation agents: the winning entry of the 2021 igibson challenge. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 1748–1755. IEEE, 2022.
  73. Nighttime single image dehazing via pixel-wise alpha blending. IEEE Access, 7:114619–114630, 2019.
  74. Multi-camera lidar inertial extension to the newer college dataset. arXiv preprint arXiv:2112.08854, 2021.
  75. Resilient robots: Concept, review, and future directions. Robotics, 6(4):22, 2017.
  76. Beyond nerf underwater: Learning neural reflectance fields for true color correction of marine imagery. IEEE Robotics and Automation Letters, 8(10):6467–6474, 2023.
  77. Go-slam: Global optimization for consistent 3d instant reconstruction. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2023.
  78. Subt-mrs: A subterranean, multi-robot, multi-spectral and multi-degraded dataset for robust slam. arXiv preprint arXiv:2307.07607, 2023.
  79. Structured3d: A large photo-realistic dataset for structured 3d modeling. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IX 16, pages 519–535. Springer, 2020.
  80. Nice-slam: Neural implicit scalable encoding for slam. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12786–12796, 2022.
  81. The uma-vi dataset: Visual–inertial odometry in low-textured and dynamic illumination environments. The International Journal of Robotics Research, 39(9):1052–1060, 2020.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (9)
  1. Xiaohao Xu (46 papers)
  2. Tianyi Zhang (262 papers)
  3. Sibo Wang (59 papers)
  4. Xiang Li (1002 papers)
  5. Yongqi Chen (8 papers)
  6. Ye Li (155 papers)
  7. Bhiksha Raj (180 papers)
  8. Matthew Johnson-Roberson (72 papers)
  9. Xiaonan Huang (32 papers)
Citations (8)
View on Semantic Scholar
Youtube Logo Streamline Icon: https://streamlinehq.com

YouTube