arXiv subjects

Chengbo Wang

Publications and source records attributed to Chengbo Wang.

At least 19 recordsLinked to original sources

arXiv2026

Rolling Shutter Camera Self-Calibration

Rolling shutter (RS) cameras are widely used in consumer devices, but their row-wise exposure causes distortions under motion, making geometric 3D vision problems dependent on both camera intrinsics and readout time ratio. Existing RS calibration methods rely on calibration targets or specialised hardware, limiting their use in unconstrained settings. We present the first self-calibration method for RS cameras that directly estimates camera intrinsics and the readout time ratio from image sequences, without requiring calibration targets. The method is implemented as a self-calibrating bundle adjustment (BA), which critically depends on the RS imaging model. We combine two known complementary models. The first formulates RS imaging as continuous-time trajectory estimation under a row-wise pose representation. The second interprets RS images as temporally distorted global shutter (GS) images and requires to estimate correction fields. The combination is non-trivial and results in a unified dual-projection model, in which each 3D point is simultaneously constrained at both row-dependent and reference timestamps along a shared continuous trajectory, enforcing stronger geometric and temporal consistency. Extensive simulations analyse the applicability of several implementations under varying conditions, and real data experiments demonstrate the accuracy, robustness, and practical effectiveness of the proposed approach.

cs.CV↗

arXiv2026

Exploration Matters for Escaping the Blur Trap in 3D Gaussian Splatting

3D Gaussian Splatting (3DGS) employs Gaussian primitives for explicit scene representation, facilitating real-time, high-fidelity reconstruction and novel view synthesis of complex scenes. However, the explicit modeling inherent in 3DGS introduces a gradient bias during optimization, rendering its non-convex optimization process highly susceptible to convergence toward local suboptimal solutions. This constitutes a fundamental limitation in 3DGS optimization, which we term the Blur Trap. To address this limitation, we integrate simple explicit exploration into the 3DGS optimization framework. First, through rigorous mathematical analysis of the 3DGS optimization formulation, we identify the underlying optimization bias responsible for the Blur Trap and categorize it into two distinct subtypes: the Far-Side Blur Trap and the Near-Side Blur Trap. Subsequently, we propose two highly straightforward exploration strategies (Random Seeding and Random Splitting) to mitigate the far-side and near-side blur traps, respectively. Experimental validation demonstrates that the incorporation of these exploration operators effectively and complementarily overcome the Blur Trap, achieving high-quality rendering performance across multiple datasets. Project page: https://chengbo-wang.github.io/ExploreGS/

cs.CV↗

arXiv2026

Global existence for quasilinear wave equations on hyperbolic space

The main purpose of this paper is to study the global solvability for a general class of quasilinear shifted wave equations on hyperbolic spaces, for smooth initial data with small amplitude. In contrast to the case of Euclidean spaces, when the space dimension is three, we do not need to assume structural conditions like the null conditions to ensure global existence. To achieve this, we establish the energy and local energy estimates for perturbed wave operators on $\mathbb{R}\times \mathbb{H}^n$. These estimates allow time-dependent metric perturbations and require only suitable smallness together with polynomial decay in the radial variable $r$. As a byproduct, for semilinear problems with power-type nonlinearities and radial data, we also obtain global solutions with low-regularity. In particular, we prove an analog of the radial Glassey conjecture on hyperbolic space.

math.AP↗

arXiv2026

Heterogeneity-Adaptive Diffusion Schrodinger Bridge for PET-Guided Whole-Body MRI Translation

While whole-body multimodal medical imaging scanners have been increasingly recognized for more effective medical applications, the excessive long acquisition time in PET-MR scanning is a major obstacle in more efficient clinical practice. Deep learning-based MRI translation provides a potential solution to reduce scan duration. However, current models often focus on specific anatomical regions and face challenges for whole-body scans that consists of highly heterogeneous feature distributions mainly due to (1) different anatomical regions across whole-body, and (2) lesions or pathological tissues. This paper tackles the challenges through a novel Heterogeneity-Adaptive Diffusion Schrodinger Bridge (HA-DSB) framework. By explicitly modeling translation as stochastic transport between source and target distributions, HA-DSB incorporates region context embeddings derived from a vision-language model (VLM) to enable region-specific modeling. To enhance fidelity of the pathological tissue, lesion-aware metabolic prior from PET is integrated directly into the bridge dynamics through a dual-stage guidance mechanism. Specifically, a PET-guided noise modulation module adaptively scales spatial diffusion perturbations during the forward process, while PET features are leveraged during the reverse process to selectively amplify lesion-relevant structures via an attention mechanism. Experiments demonstrate the superiority of our method across different body regions in whole-body MRI translation and show improved translation quality in lesion areas under PET guidance. Our code is available at Github.

cs.CV↗

arXiv2025

DisorientLiDAR: Physical Attacks on LiDAR-based Localization

Deep learning models have been shown to be susceptible to adversarial attacks with visually imperceptible perturbations. Even this poses a serious security challenge for the localization of self-driving cars, there has been very little exploration of attack on it, as most of adversarial attacks have been applied to 3D perception. In this work, we propose a novel adversarial attack framework called DisorientLiDAR targeting LiDAR-based localization. By reverse-engineering localization models (e.g., feature extraction networks), adversaries can identify critical keypoints and strategically remove them, thereby disrupting LiDAR-based localization. Our proposal is first evaluated on three state-of-the-art point-cloud registration models (HRegNet, D3Feat, and GeoTransformer) using the KITTI dataset. Experimental results demonstrate that removing regions containing Top-K keypoints significantly degrades their registration accuracy. We further validate the attack's impact on the Autoware autonomous driving platform, where hiding merely a few critical regions induces noticeable localization drift. Finally, we extended our attacks to the physical world by hiding critical regions with near-infrared absorptive materials, thereby successfully replicate the attack effects observed in KITTI data. This step has been closer toward the realistic physical-world attack that demonstrate the veracity and generality of our proposal.

cs.CV↗

arXiv2025

PIS3R: Very Large Parallax Image Stitching via Deep 3D Reconstruction

Image stitching aim to align two images taken from different viewpoints into one seamless, wider image. However, when the 3D scene contains depth variations and the camera baseline is significant, noticeable parallax occurs-meaning the relative positions of scene elements differ substantially between views. Most existing stitching methods struggle to handle such images with large parallax effectively. To address this challenge, in this paper, we propose an image stitching solution called PIS3R that is robust to very large parallax based on the novel concept of deep 3D reconstruction. First, we apply visual geometry grounded transformer to two input images with very large parallax to obtain both intrinsic and extrinsic parameters, as well as the dense 3D scene reconstruction. Subsequently, we reproject reconstructed dense point cloud onto a designated reference view using the recovered camera parameters, achieving pixel-wise alignment and generating an initial stitched image. Finally, to further address potential artifacts such as holes or noise in the initial stitching, we propose a point-conditioned image diffusion module to obtain the refined result.Compared with existing methods, our solution is very large parallax tolerant and also provides results that fully preserve the geometric integrity of all pixels in the 3D photogrammetric context, enabling direct applicability to downstream 3D vision tasks such as SfM. Experimental results demonstrate that the proposed algorithm provides accurate stitching results for images with very large parallax, and outperforms the existing methods qualitatively and quantitatively.

cs.CV↗

arXiv2025

DUN-SRE: Deep Unrolling Network with Spatiotemporal Rotation Equivariance for Dynamic MRI Reconstruction

Dynamic Magnetic Resonance Imaging (MRI) exhibits transformation symmetries, including spatial rotation symmetry within individual frames and temporal symmetry along the time dimension. Explicit incorporation of these symmetry priors in the reconstruction model can significantly improve image quality, especially under aggressive undersampling scenarios. Recently, Equivariant convolutional neural network (ECNN) has shown great promise in exploiting spatial symmetry priors. However, existing ECNNs critically fail to model temporal symmetry, arguably the most universal and informative structural prior in dynamic MRI reconstruction. To tackle this issue, we propose a novel Deep Unrolling Network with Spatiotemporal Rotation Equivariance (DUN-SRE) for Dynamic MRI Reconstruction. The DUN-SRE establishes spatiotemporal equivariance through a (2+1)D equivariant convolutional architecture. In particular, it integrates both the data consistency and proximal mapping module into a unified deep unrolling framework. This architecture ensures rigorous propagation of spatiotemporal rotation symmetry constraints throughout the reconstruction process, enabling more physically accurate modeling of cardiac motion dynamics in cine MRI. In addition, a high-fidelity group filter parameterization mechanism is developed to maintain representation precision while enforcing symmetry constraints. Comprehensive experiments on Cardiac CINE MRI datasets demonstrate that DUN-SRE achieves state-of-the-art performance, particularly in preserving rotation-symmetric structures, offering strong generalization capability to a broad range of dynamic MRI reconstruction tasks.

eess.IV↗

arXiv2024

Faster and Better 3D Splatting via Group Training

3D Gaussian Splatting (3DGS) has emerged as a powerful technique for novel view synthesis, demonstrating remarkable capability in high-fidelity scene reconstruction through its Gaussian primitive representations. However, the computational overhead induced by the massive number of primitives poses a significant bottleneck to training efficiency. To overcome this challenge, we propose Group Training, a simple yet effective strategy that organizes Gaussian primitives into manageable groups, optimizing training efficiency and improving rendering quality. This approach shows universal compatibility with existing 3DGS frameworks, including vanilla 3DGS and Mip-Splatting, consistently achieving accelerated training while maintaining superior synthesis quality. Extensive experiments reveal that our straightforward Group Training strategy achieves up to 30\% faster convergence and improved rendering quality across diverse scenarios. Project Website: https://chengbo-wang.github.io/3DGS-with-Group-Training/

cs.CV↗

arXiv2024

SRE-CNN: A Spatiotemporal Rotation-Equivariant CNN for Cardiac Cine MR Imaging

Dynamic MR images possess various transformation symmetries,including the rotation symmetry of local features within the image and along the temporal dimension. Utilizing these symmetries as prior knowledge can facilitate dynamic MR imaging with high spatiotemporal resolution. Equivariant CNN is an effective tool to leverage the symmetry priors. However, current equivariant CNN methods fail to fully exploit these symmetry priors in dynamic MR imaging. In this work, we propose a novel framework of Spatiotemporal Rotation-Equivariant CNN (SRE-CNN), spanning from the underlying high-precision filter design to the construction of the temporal-equivariant convolutional module and imaging model, to fully harness the rotation symmetries inherent in dynamic MR images. The temporal-equivariant convolutional module enables exploitation the rotation symmetries in both spatial and temporal dimensions, while the high-precision convolutional filter, based on parametrization strategy, enhances the utilization of rotation symmetry of local features to improve the reconstruction of detailed anatomical structures. Experiments conducted on highly undersampled dynamic cardiac cine data (up to 20X) have demonstrated the superior performance of our proposed approach, both quantitatively and qualitatively.

eess.IV↗

arXiv2024

Blow-up of solutions to semilinear wave equations with spatial derivatives

For small-amplitude semilinear wave equations with power type nonlinearity on the first-order spatial derivative, the expected sharp upper bound on the lifespan of solutions is obtained for both critical cases and subcritical cases, for all spatial dimensions $n>1$. It is achieved uniformly by constructing the integral equations, deriving the ordinary differential inequality system, and iteration argument. Combined with the former works, the sharp lifespan estimates for this problem are completely established, at least for the spherical symmetric case.

math.AP↗

arXiv2024

Critical conditions of nonlinearities in the John problem

In this paper, we determine the sharp criteria for the nonlinearities in the John problem $\partial_{t}^2 u-\Delta_{\R^3}u= F(u)$ so that the problem admits global solutions for small initial data with compact support. The criteria is of Dini type. For the situation in which the criteria is not satisfied, we show that the solution will blow up in finite time for some small initial data, together with an almost sharp estimates of the lifespan. For the proof, we mainly rely on the precise pointwise estimates of the solution, based on the fundamental solution, John's iteration argument as well as an adaption of the slicing method.

math.AP↗

arXiv2023

Wave equations with logarithmic nonlinearity on hyperbolic spaces

In light of the exponential decay of solutions of linear wave equations on hyperbolic spaces $\mathbb{H}^n$, to illustrate the critical nature, we investigate nonlinear wave equations with logarithmic nonlinearity, which behaves like $\left(\ln {1}/{|u|}\right)^{1-p}|u|$ near $u=0$, on hyperbolic spaces. Concerning the global existence vs blow up with small data, we expect that the problem admits a critical power $p_c(n)>1$. When $n=3$, we prove that the critical power is $3$, by proving global existence for $p>3$, as well as generically blow up for $p\in (1,3)$.

math.AP↗

arXiv2022

Blow-up for semilinear wave equations on Kerr black hole backgrounds

We examine solutions to semilinear wave equations on black hole backgrounds and give a proof of an analog of the blow up part of the John theorem, with $F_p(u)=|u|^{p}$, on the Schwarzschild and Kerr black hole backgrounds. Concerning the case of Schwarzschild, we construct a class of small data, so that the solution blows up along the outgoing null cone, which applies for both $F_p(u)=|u|^{p}$ and the focusing nonlinearity $F_p(u)=|u|^{p-1}u$. The proof suggests that the black hole does not have any essential influence on the formation of singularity, in the region away from the Cauchy horizon $r=r_-$ or the singularity $r=0$. Our approach is also robust enough to be adapted for general asymptotically flat space-time manifolds, possibly exterior to a compact domain, with spatial dimension $n\ge 2$. Typical examples include exterior domains, asymptotically Euclidean spaces, Reissner-Nördström space-times, and Kerr-Newman space-times.

math.AP↗

arXiv2022

Global Solutions with Small Initial Data to Semilinear Wave Equations with Energy Supercritical Powers

Considering $1+n$ dimensional semilinear wave equations with energy supercritical powers $p> 1+4/(n-2)$, we obtain global solutions for any initial data with small norm in $H^{s_c}\times H^{s_c-1}$, under the technical smooth condition $p>s_c-\bar{s}_0$, with $\bar{s}_0= 1/2+(n-3)/(2\max(n-1-p,n-3))$ and $s_c=n/2-2/(p-1)$. In particular, combined with previous works, our results give a complete verification of the Strauss conjecture, up to space dimension $9$. The higher dimensional case, $n\ge 10$, seems to be unreachable, in view of the wellposed theory in $H^s$.

math.AP↗

arXiv2022

Deep Residual Shrinkage Networks for EMG-based Gesture Identification

This work introduces a method for high-accuracy EMG based gesture identification. A newly developed deep learning method, namely, deep residual shrinkage network is applied to perform gesture identification. Based on the feature of EMG signal resulting from gestures, optimizations are made to improve the identification accuracy. Finally, three different algorithms are applied to compare the accuracy of EMG signal recognition with that of DRSN. The result shows that DRSN excel traditional neural networks in terms of EMG recognition accuracy. This paper provides a reliable way to classify EMG signals, as well as exploring possible applications of DRSN.

eess.SP↗

arXiv2022

An alternative proof of Tataru's dispersive estimates

The aim of this article is to give an alternative proof of Tataru's dispersive estimates for wave equations posed on the hyperbolic space. Based on the formula for the wave kernel on hyperbolic spaces, in \cite{MR2743652}, we give the proof from the perspective of Bessel potentials, by exploiting various facts about Gamma functions, modified Bessel functions, and Bessel potentials. This leads to our proof being more self-contained than that in Tataru \cite{MR1804518}.

math.AP↗

arXiv2021

Long time existence for semilinear wave equations with the inverse-square potential

In this paper, we study the semilinear wave equations with the inverse-square potential. By transferring the original equation to a "fractional dimensional" wave equation and analyzing the properties of its fundamental solution, we establish a long-time existence result, for sufficiently small, spherically symmetric initial data. Together with the previously known blow-up result, we determine the critical exponent which divides the global existence and finite time blow-up. Moreover, the sharp lower bounds of the lifespan are obtained, except for certain borderline case. In addition, our technology allows us to handle an extreme case for the potential, which has hardly been discussed in literature.

math.AP↗

arXiv2021

Reversed Strichartz estimates for wave on non-trapping asymptotically hyperbolic manifolds and applications

We provide reversed Strichartz estimates for the shifted wave equations on non-trapping asymptotically hyperbolic manifolds using cluster estimates for spectral projectors proved previously in such generality. As a consequence, we solve a problem left open in \cite{SSWZ} about the endpoint case for global well-posedness of nonlinear wave equations. We also provide estimates in this context for the maximal wave operator.

math.AP↗