pith. machine review for the scientific record. sign in

arxiv: 2604.16540 · v1 · submitted 2026-04-17 · 💻 cs.CV · cs.AI

Recognition: unknown

PoInit-of-View: Poisoning Initialization of Views Transfers Across Multiple 3D Reconstruction Systems

Weijie Wang , Songlong Xing , Zhengyu Zhao , Nicu Sebe , Bruno Lepri
Authors on Pith no claims yet

Pith reviewed 2026-05-10 08:53 UTC · model grok-4.3

classification 💻 cs.CV cs.AI
keywords reconstructionsystemspoinit-of-viewpoisoningviewsacrossadversarialcross-view
0
0 comments X

The pith

PoInit-of-View poisons SfM initialization by optimizing cross-view gradient inconsistencies to disrupt keypoint detection and feature matching, yielding transferable degradation in rendered 3D reconstruction quality across systems.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

3D reconstruction systems often start by estimating camera positions and 3D points from a set of input images using structure-from-motion. The attack adds small, carefully chosen noise to those images so that the same 3D point projects to locations whose gradients point in inconsistent directions across views. This breaks the keypoint detectors and matchers that SfM relies on, producing wrong camera poses and bad 3D point clouds. The corrupted initialization then propagates through later stages such as NeRF or 3DGS training, resulting in blurry or distorted output views. Experiments show the poisoned views transfer better than simply attacking a single reconstruction pipeline end-to-end.

Core claim

we propose PoInit-of-View, which optimizes adversarial perturbations to intentionally introduce cross-view gradient inconsistencies at projections of corresponding 3D points. These inconsistencies disrupt keypoint detection and feature matching, thereby corrupting pose estimation and triangulation within SfM, eventually resulting in low-quality rendered views.

Load-bearing premise

That cross-view gradient inconsistencies optimized on one SfM implementation will reliably collapse correspondences in the SfM modules of unseen target reconstruction systems without any access to those systems' parameters or training data.

Figures

Figures reproduced from arXiv: 2604.16540 by Bruno Lepri, Nicu Sebe, Songlong Xing, Weijie Wang, Zhengyu Zhao.

Figure 1
Figure 1. Figure 1: Our PoInit-of-View method injects imperceptible pertur￾bations into input views (top) to induce cross-view inconsistency (see details in [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Framework of PoInit-of-View. for subsequent dense reconstruction or neural optimization; without a reliable SfM initialization, downstream optimiz￾ers often fail to converge to a coherent scene. Formally, let {Ii} N i=1 denote the input views. For image i, we denote keypoints by pi,k ∈ R 2 and their associated descriptors by ϕ(pi,k) ∈ R d . Candidate correspondences between a keypoint pi,k in image i and p… view at source ↗
Figure 3
Figure 3. Figure 3: Cross-view gradient inconsistency causes geometric [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Qualitative comparison of clean and poisoned reconstructions across multiple scenes on [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: visualizes the clean, poisoned, and amplified￾difference views to qualitatively assess the subtlety of Clean Poisoned Amplified Diff ( ×8) [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: Gradient-difference comparison on T&T. Our perturba￾tion remains imperceptible while inducing clear edge-aligned dis￾crepancies, unlike random noise, indicating targeted disruption of SfM-relevant structures. 0.0 0.2 0.26 0.4 0.6 0.8 1.0 1.1 Weight of Lth 40 60 80 Registered images (%) Lth = g + d/ r 20 40 60 80 Triangulated keypoints (k) Registered images (%) Triangulated keypoints (k) The estimated Lth … view at source ↗
Figure 9
Figure 9. Figure 9: Impact of cross-view inconsistency on SfM. Increasing [PITH_FULL_IMAGE:figures/full_fig_p008_9.png] view at source ↗
Figure 7
Figure 7. Figure 7: Ablation on perturbation budget ε, showing that larger perturbations cause greater degradation in registration and recon￾struction. Effect of Structured Perturbations vs. Random Noise. We compute gradient-difference maps using the Sobel [14] operator between the clean and poisoned images for visual￾ization, as shown in [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
read the original abstract

Poisoning input views of 3D reconstruction systems has been recently studied. However, we identify that existing studies simply backpropagate adversarial gradients through the 3D reconstruction pipeline as a whole, without uncovering the new vulnerability rooted in specific modules of the 3D reconstruction pipeline. In this paper, we argue that the structure-from-motion (SfM) initialization, as the geometric core of many widely used reconstruction systems, can be targeted to achieve transferable poisoning effects across diverse 3D reconstruction systems. To this end, we propose PoInit-of-View, which optimizes adversarial perturbations to intentionally introduce cross-view gradient inconsistencies at projections of corresponding 3D points. These inconsistencies disrupt keypoint detection and feature matching, thereby corrupting pose estimation and triangulation within SfM, eventually resulting in low-quality rendered views. We also provide a theoretical analysis that connects cross-view inconsistency to correspondence collapse. Experimental results demonstrate the effectiveness of our PoInit-of-View on diverse 3D reconstruction systems and datasets, surpassing the single-view baseline by 25.1% in PSNR and 16.5% in SSIM in black-box transfer settings, such as 3DGS to NeRF.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review supplies insufficient detail to enumerate concrete free parameters, axioms, or invented entities; the method implicitly assumes standard differentiability of SfM components and the existence of corresponding 3D points across views.

axioms (1)
  • domain assumption SfM pipelines rely on keypoint detection and feature matching that are sensitive to small gradient inconsistencies at projected 3D points
    Invoked to justify why the optimized perturbations corrupt pose estimation and triangulation

pith-pipeline@v0.9.0 · 5526 in / 1346 out tokens · 21496 ms · 2026-05-10T08:53:39.389342+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Medoid Prototype Alignment for Cross-Plant Unknown Attack Detection in Industrial Control Systems

    cs.CR 2026-04 unverdicted novelty 6.0

    Medoid prototype alignment detects unknown attacks across industrial plants by aligning domain-specific medoid summaries rather than raw samples, yielding 0.843 average accuracy on gas and water system transfers.

Reference graph

Works this paper leans on

58 extracted references · 2 canonical work pages · cited by 1 Pith paper · 1 internal anchor

  1. [1]

    Barron, Ben Mildenhall, Matthew Tancik, Pe- ter Hedman, Ricardo Martin-Brualla, and Pratul P

    Jonathan T. Barron, Ben Mildenhall, Matthew Tancik, Pe- ter Hedman, Ricardo Martin-Brualla, and Pratul P. Srini- vasan. Mip-nerf 360: Unbounded anti-aliased neural radi- ance fields. InCVPR, 2022. 1, 5

    2022

  2. [2]

    Sg-nerf: Neural surface recon- struction with scene graph optimization

    Yiyang Chen, Siyan Dong, Xulong Wang, Lulu Cai, Youyi Zheng, and Yanchao Yang. Sg-nerf: Neural surface recon- struction with scene graph optimization. InECCV, 2024. 1, 2

    2024

  3. [3]

    Boosting adversarial at- tacks with momentum

    Yinpeng Dong, Fangzhou Liao, Tianyu Pang, Hang Su, Jun Zhu, Xiaolin Hu, and Jianguo Li. Boosting adversarial at- tacks with momentum. InCVPR, 2018. 2

    2018

  4. [4]

    Viewfool: Evaluating the robustness of visual recognition to adversarial viewpoints

    Yinpeng Dong, Shouwei Ruan, Hang Su, Caixin Kang, Xingxing Wei, and Jun Zhu. Viewfool: Evaluating the robustness of visual recognition to adversarial viewpoints. NeurIPS, 2022. 2

    2022

  5. [5]

    Gausstrap: Stealthy poisoning attacks on 3d gaussian splat- ting for targeted scene confusion, 2025

    Jiaxin Hong, Sixu Chen, Shuoyang Sun, Hongyao Yu, Hao Fang, Yuqi Tan, Bin Chen, Shuhan Qi, and Jiawei Li. Gausstrap: Stealthy poisoning attacks on 3d gaussian splat- ting for targeted scene confusion, 2025. 2

    2025

  6. [6]

    Image quality metrics: Psnr vs

    Alain Hore and Djamel Ziou. Image quality metrics: Psnr vs. ssim.ICPR, pages 2366–2369, 2010. 5, 7

    2010

  7. [7]

    Targeted adversarial at- tacks on generalizable neural radiance fields

    Andr ´as Horv´ath and Csaba M J´ozsa. Targeted adversarial at- tacks on generalizable neural radiance fields. InICCV, 2023. 1, 2

    2023

  8. [8]

    Driveadapter: Breaking the coupling barrier of perception and planning in end-to-end autonomous driving

    Xiaosong Jia, Yulu Gao, Li Chen, Junchi Yan, Patrick Langechuan Liu, and Hongyang Li. Driveadapter: Breaking the coupling barrier of perception and planning in end-to-end autonomous driving. InICCV), 2023. 1

    2023

  9. [9]

    Vad: Vectorized scene representation for efficient autonomous driving

    Bo Jiang, Shaoyu Chen, Qing Xu, Bencheng Liao, Jiajie Chen, Helong Zhou, Qian Zhang, Wenyu Liu, Chang Huang, and Xinggang Wang. Vad: Vectorized scene representation for efficient autonomous driving. InICCV, 2023. 1

    2023

  10. [10]

    Nerfail: Neural radiance fields-based multi- view adversarial attack

    Wenxiang Jiang, Hanwei Zhang, Xi Wang, Zhongwen Guo, and Hao Wang. Nerfail: Neural radiance fields-based multi- view adversarial attack. InAAAI, 2024. 2

    2024

  11. [11]

    Mpam-3dgs: Multi- parametric adversarial manipulation for 3d gaussian splat- ting

    Wenxiang Jiang, Hanwei Zhang, Weigang Wang, Zhongwen Guo, Tianao Zhang, and Hao Wang. Mpam-3dgs: Multi- parametric adversarial manipulation for 3d gaussian splat- ting. InICASSP. IEEE, 2025. 2

    2025

  12. [12]

    Stealthattack: Robust 3d gaussian splatting poisoning via density-guided illusions

    Bo-Hsu Ke, You-Zhe Xie, Yu-Lun Liu, and Wei-Chen Chiu. Stealthattack: Robust 3d gaussian splatting poisoning via density-guided illusions. InICCV, 2025. 1, 2

    2025

  13. [13]

    3d gaussian splatting for real-time radiance field rendering.ACM Transactions on Graphics,

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering.ACM Transactions on Graphics,

  14. [14]

    On the accuracy of the sobel edge detector

    Josef Kittler. On the accuracy of the sobel edge detector. Image and Vision Computing, 1983. 8

    1983

  15. [15]

    Tanks and temples: Benchmarking large-scale scene reconstruction

    Arno Knapitsch, Jaesik Park, Qian-Yi Zhou, and Vladlen Koltun. Tanks and temples: Benchmarking large-scale scene reconstruction. InACM Transactions on Graphics (TOG),

  16. [16]

    NerfBaselines: Consis- tent and Reproducible Evaluation of Novel View Synthesis Methods

    Jonas Kulhanek and Torsten Sattler. NerfBaselines: Consis- tent and Reproducible Evaluation of Novel View Synthesis Methods. InNeurIPS, 2025. 6

    2025

  17. [17]

    Freeinsert: Disentangled text-guided object insertion in 3d gaussian scene without spatial priors, 2025

    Chenxi Li, Weijie Wang, Qiang Li, Bruno Lepri, Nicu Sebe, and Weizhi Nie. Freeinsert: Disentangled text-guided object insertion in 3d gaussian scene without spatial priors, 2025. 1

    2025

  18. [18]

    Cross-modal and uncertainty-aware agglomeration for open- vocabulary 3d scene understanding

    Jinlong Li, Cristiano Saltori, Fabio Poiesi, and Nicu Sebe. Cross-modal and uncertainty-aware agglomeration for open- vocabulary 3d scene understanding. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19390–19400, 2025

    2025

  19. [19]

    Token Reduction via Local and Global Contexts Optimization for Efficient Video Large Language Models

    Jinlong Li, Liyuan Jiang, Haonan Zhang, and Nicu Sebe. Token reduction via local and global contexts optimization for efficient video large language models.arXiv preprint arXiv:2603.01400, 2026. 1

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  20. [20]

    Adv3d: Generating 3d adversarial examples in driving sce- narios with nerf

    Leheng Li, Yiming Zhang, Jialei Wang, and Aimin Zhou. Adv3d: Generating 3d adversarial examples in driving sce- narios with nerf. InNeurIPS, 2023. 2

    2023

  21. [21]

    Robotic visual instruction

    Yanbang Li, Ziyang Gong, Haoyang Li, Xiaoqi Huang, Haolan Kang, Guangping Bai, and Xianzheng Ma. Robotic visual instruction. InCVPR, 2025. 1

    2025

  22. [22]

    Barf: Bundle-adjusting neural radiance fields

    Chen-Hsuan Lin, Wei-Chiu Ma, Antonio Torralba, and Si- mon Lucey. Barf: Bundle-adjusting neural radiance fields. InICCV, 2021. 2

    2021

  23. [23]

    Nerfool: Uncovering the vulnerability of generaliz- able neural radiance fields against adversarial perturbations

    Yingyan Lin, Kai Zhang, Meng Wang, Zhiyuan Li, and Fan Yang. Nerfool: Uncovering the vulnerability of generaliz- able neural radiance fields against adversarial perturbations. InICML, 2023. 1, 2, 6

    2023

  24. [24]

    Delving into transferable adversarial examples and black- box attacks

    Yanpei Liu, Xinyun Chen, Chang Liu, and Dawn Song. Delving into transferable adversarial examples and black- box attacks. InICLR, 2017. 2

    2017

  25. [25]

    Distinctive image features from scale- invariant keypoints

    David G Lowe. Distinctive image features from scale- invariant keypoints. InIJCV, 2004. 4

    2004

  26. [26]

    Poison-splat: Computation cost attack on 3d gaussian splatting

    Jiahao Lu, Yifan Zhang, Qiuhong Shen, Xinchao Wang, and Shuicheng Y AN. Poison-splat: Computation cost attack on 3d gaussian splatting. InICLR, 2025. 2

    2025

  27. [27]

    Real-time simulated avatar from head-mounted sensors

    Zhengyi Luo, Jinkun Cao, Rawal Khirodkar, Alexander Win- kler, Kris Kitani, and Weipeng Xu. Real-time simulated avatar from head-mounted sensors. InCVPR, 2024. 1

    2024

  28. [28]

    Il2-nerf: Advancing adversarial robustness in generalizable nerfs

    Nicole Meng, Rui Zhao, Xin Wang, and Hao Zhou. Il2-nerf: Advancing adversarial robustness in generalizable nerfs. In CVPR, 2025. 1, 2, 6

    2025

  29. [29]

    Nerf: Representing scenes as neural radiance fields for view syn- thesis.Communications of the ACM, 2021

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view syn- thesis.Communications of the ACM, 2021. 2, 5

    2021

  30. [30]

    Instant neural graphics primitives with a multires- olution hash encoding.ACM Trans

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. Instant neural graphics primitives with a multires- olution hash encoding.ACM Trans. Graph., 41(4):102:1– 102:15, 2022. 2, 5, 7

    2022

  31. [31]

    T2td: Text-3d generation model based on prior knowledge guidance.IEEE TPAMI, 2024

    Weizhi Nie, Ruidong Chen, Weijie Wang, Bruno Lepri, and Nicu Sebe. T2td: Text-3d generation model based on prior knowledge guidance.IEEE TPAMI, 2024. 1

    2024

  32. [32]

    Rethinking depth estimation for multi- view stereo: A unified representation

    Rui Peng, Rongjie Wang, Zhenyu Wang, Yawen Lai, and Ronggang Wang. Rethinking depth estimation for multi- view stereo: A unified representation. InCVPR, 2022. 2

    2022

  33. [33]

    Unmix-nerf: Spectral un- mixing meets neural radiance fields.ICCV, 2025

    Fabian Perez, Sara Rojas, Carlos Hinojosa, Hoover Rueda- Chac ˜A`gn, and Bernard Ghanem. Unmix-nerf: Spectral un- mixing meets neural radiance fields.ICCV, 2025. 2

    2025

  34. [34]

    Structure-from-motion revisited

    Johannes Lutz Sch ¨onberger and Jan-Michael Frahm. Structure-from-motion revisited. InCVPR, 2016. 2, 5

    2016

  35. [35]

    Pixelwise view selection for unstructured multi-view stereo

    Johannes L Sch ¨onberger, Enliang Zheng, Jan-Michael Frahm, and Marc Pollefeys. Pixelwise view selection for unstructured multi-view stereo. InECCV, 2016. 2, 5

    2016

  36. [36]

    Deep- voxels: Learning persistent 3d feature embeddings

    Vincent Sitzmann, Justus Thies, Felix Heide, Matthias Nießner, Gordon Wetzstein, and Michael Zollh ¨ofer. Deep- voxels: Learning persistent 3d feature embeddings. In CVPR, 2019. 2

    2019

  37. [37]

    Geometry cloak: Preventing tgs-based 3d re- construction from copyrighted images.NeurIPS, 2024

    Qi Song, Ziyuan Luo, Ka Chun Cheung, Simon See, and Renjie Wan. Geometry cloak: Preventing tgs-based 3d re- construction from copyrighted images.NeurIPS, 2024. 1, 2

    2024

  38. [38]

    Neuralrecon: Real-time coherent 3d reconstruc- tion from monocular video

    Jiaming Sun, Yiming Xie, Linghao Chen, Xiaowei Zhou, and Hujun Bao. Neuralrecon: Real-time coherent 3d reconstruc- tion from monocular video. InCVPR, 2021. 2

    2021

  39. [39]

    Vgos: V oxel grid op- timization for view synthesis from sparse inputs

    Jiakai Sun, Zhanjie Zhang, Jiafu Chen, Guangyuan Li, Boyan Ji, Lei Zhao, and Wei Xing. Vgos: V oxel grid op- timization for view synthesis from sparse inputs. InIJCAI,

  40. [40]

    S. Ullman. The interpretation of structure from motion.Pro- ceedings of the Royal Society of London. Series B. Biological Sciences, 203(1153):405–426, 1979. 1

    1979

  41. [41]

    Benchmarking robustness in neural radiance fields

    Chen Wang, Angtian Wang, Junbo Li, Alan Yuille, and Cihang Xie. Benchmarking robustness in neural radiance fields. InCVPR, 2024. 1

    2024

  42. [42]

    Turn fake into real: Adversarial head turn attacks against deepfake detection

    Weijie Wang, Zhengyu Zhao, Nicu Sebe, and Bruno Lepri. Turn fake into real: Adversarial head turn attacks against deepfake detection.arXiv preprint arXiv:2309.01104, 2023. 1

    work page arXiv 2023

  43. [43]

    Uvmap-id: A controllable and personalized uv map generative model

    Weijie Wang, Jichao Zhang, Chang Liu, Xia Li, Xingqian Xu, Humphrey Shi, Nicu Sebe, and Bruno Lepri. Uvmap-id: A controllable and personalized uv map generative model. In ACM MM, pages 10725–10734, 2024. 1

    2024

  44. [44]

    Fully-geometric cross-attention for point cloud registration

    Weijie Wang, Guofeng Mei, Jian Zhang, Nicu Sebe, Bruno Lepri, and Fabio Poiesi. Fully-geometric cross-attention for point cloud registration. In3DV. IEEE, 2025. 1

    2025

  45. [45]

    Mvster: Epipo- lar transformer for efficient multi-view stereo

    Xiaofeng Wang, Zheng Zhu, Guan Huang, Fangbo Qin, Yun Ye, Yijia He, Xu Chi, and Xingang Wang. Mvster: Epipo- lar transformer for efficient multi-view stereo. InECCV. Springer, 2022. 2

    2022

  46. [46]

    Wild-gs: Real- time novel view synthesis from unconstrained photo collec- tions.NeurIPS, 2024

    Jiacong Xu, Yiqun Mei, and Vishal Patel. Wild-gs: Real- time novel view synthesis from unconstrained photo collec- tions.NeurIPS, 2024. 2

    2024

  47. [47]

    Imvid: Immersive volumetric videos for en- hanced vr engagement

    Zhengxian Yang, Shi Pan, Shengqi Wang, Haoxiang Wang, Li Lin, Guanjun Li, Zhengqi Wen, Borong Lin, Jianhua Tao, and Tao Yu. Imvid: Immersive volumetric videos for en- hanced vr engagement. InCVPR, 2025. 1

    2025

  48. [48]

    Mvsnet: Depth inference for unstructured multi-view stereo

    Yao Yao, Zixin Luo, Shiwei Li, Tian Fang, and Long Quan. Mvsnet: Depth inference for unstructured multi-view stereo. InECCV, 2018. 2

    2018

  49. [49]

    Fewviewgs: Gaussian splatting with few view matching and multi-stage training.NeurIPS, 2024

    Ruihong Yin, Vladimir Yugay, Yue Li, Sezer Karaoglu, and Theo Gevers. Fewviewgs: Gaussian splatting with few view matching and multi-stage training.NeurIPS, 2024. 2

    2024

  50. [50]

    Mip-splatting: Alias-free 3d gaussian splat- ting.CVPR, 2024

    Zehao Yu, Anpei Chen, Binbin Huang, Torsten Sattler, and Andreas Geiger. Mip-splatting: Alias-free 3d gaussian splat- ting.CVPR, 2024. 5, 6, 7, 8

    2024

  51. [51]

    Vladimir Yugay, Theo Gevers, and Martin R. Oswald. Magic-slam: Multi-agent gaussian globally consistent slam. InCVPR, 2025. 1

    2025

  52. [52]

    Gaussian splatting under attack: Investigating ad- versarial noise in 3d objects

    Abdurrahman Zeybey, Mehmet Ergezer, and Tommy Nguyen. Gaussian splatting under attack: Investigating ad- versarial noise in 3d objects. InNeurIPS, 2024. 2

    2024

  53. [53]

    Structural multiplane image: Bridging neural view synthesis and 3d reconstruction

    Mingfang Zhang, Jinglu Wang, Xiao Li, Yifei Huang, Yoichi Sato, and Yan Lu. Structural multiplane image: Bridging neural view synthesis and 3d reconstruction. InCVPR, 2023. 2

    2023

  54. [54]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. InCVPR, 2018. 5, 7

    2018

  55. [55]

    On suc- cess and simplicity: A second look at transferable targeted attacks.NeurIPS, 2021

    Zhengyu Zhao, Zhuoran Liu, and Martha Larson. On suc- cess and simplicity: A second look at transferable targeted attacks.NeurIPS, 2021. 2

    2021

  56. [56]

    Revisiting transferable adversarial images: System- ization, evaluation, and new insights.IEEE Transactions on Pattern Analysis and Machine Intelligence, pages 1–16,

    Zhengyu Zhao, Hanwei Zhang, Renjue Li, Ronan Sicre, Lau- rent Amsaleg, Michael Backes, Qi Li, Qian Wang, and Chao Shen. Revisiting transferable adversarial images: System- ization, evaluation, and new insights.IEEE Transactions on Pattern Analysis and Machine Intelligence, pages 1–16,

  57. [57]

    Physical 3d adversarial attacks against monocular depth estimation in autonomous driving

    Junhao Zheng, Chenhao Lin, Jiahao Sun, Zhengyu Zhao, Qian Li, and Chao Shen. Physical 3d adversarial attacks against monocular depth estimation in autonomous driving. InCVPR, 2024. 2

    2024

  58. [58]

    Stereo magnification: Learning view syn- thesis using multiplane images

    Tinghui Zhou, Richard Tucker, John Flynn, Graham Fyffe, and Noah Snavely. Stereo magnification: Learning view syn- thesis using multiplane images. InSIGGRAPH Asia, 2018. 2

    2018