Approach For Surface Reconstruction Research Articles

Gaussian Opacity Fields: Efficient Adaptive Surface Reconstruction in Unbounded Scenes

Recently, 3D Gaussian Splatting (3DGS) has demonstrated impressive novel view synthesis results, while allowing the rendering of high-resolution images in real-time. However, leveraging 3D Gaussians for surface reconstruction poses significant challenges due to the explicit and disconnected nature of 3D Gaussians. In this work, we present Gaussian Opacity Fields (GOF), a novel approach for efficient, high-quality, and adaptive surface reconstruction in unbounded scenes. Our GOF is derived from ray-tracing-based volume rendering of 3D Gaussians, enabling direct geometry extraction from 3D Gaussians by identifying its levelset, without resorting to Poisson reconstruction or TSDF fusion as in previous work. We approximate the surface normal of Gaussians as the normal of the ray-Gaussian intersection plane, enabling the application of regularization that significantly enhances geometry. Furthermore, we develop an efficient geometry extraction method utilizing Marching Tetrahedra, where the tetrahedral grids are induced from 3D Gaussians and thus adapt to the scene's complexity. Our evaluations reveal that GOF surpasses existing 3DGS-based methods in surface reconstruction and novel view synthesis. Further, it compares favorably to or even outperforms, neural implicit methods in both quality and speed.

Shadow-constrained shape-from-shading for pixel-wise 3D surface reconstruction at the lunar south pole

ABSTRACT High-resolution digital elevation models (DEMs) of the lunar surface are crucial for lunar exploration missions and scientific research. The emergence of advanced technologies, such as Shape-from-Shading (SfS), has enabled the generation of pixel-wise high-resolution DEMs from monocular images of the lunar surface. However, SfS encounters significant challenges in locations with limited illumination, such as the lunar south pole, where the surface is largely covered by shadows. Therefore, this paper presents a novel shadow-constrained SfS approach for pixel-wise 3D surface reconstruction at the lunar south pole. The proposed approach uses multiple high-resolution images captured under different illumination conditions and an existing low-resolution DEM as the inputs and generates a high-resolution DEM with the same resolution as that of the input image through hierarchical SfS processing incorporating shadow constraints. Experiments were conducted using actual images collected by the Lunar Reconnaissance Orbiter (LRO) Narrow Angle Camera (NAC) at the lunar south pole. Comparisons with respect to photogrammetric DEMs generated from stereo NAC images show that the DEMs generated using the proposed approach exhibit the smallest root mean square error (RMSE). Moreover, shaded relief images rendered from the DEMs generated using the proposed approach demonstrate the highest similarity to the actual NAC images. Detailed profile comparisons further validate the effectiveness of the shadow constraint in optimizing 3D reconstruction in regions in proximity to shadows and within shadowed regions. The proposed shadow-constrained SfS approach can be used to generate high-resolution DEMs to support future missions to explore the lunar south pole, with applications including landing site evaluation and route planning for lunar probes or astronauts.

SC-NeuS: Consistent Neural Surface Reconstruction from Sparse and Noisy Views

The recent neural surface reconstruction approaches using volume rendering have made much progress by achieving impressive surface reconstruction quality, but are still limited to dense and highly accurate posed views. To overcome such drawbacks, this paper pays special attention on the consistent surface reconstruction from sparse views with noisy camera poses. Unlike previous approaches, the key difference of this paper is to exploit the multi-view constraints directly from the explicit geometry of the neural surface, which can be used as effective regularization to jointly learn the neural surface and refine the camera poses. To build effective multi-view constraints, we introduce a fast differentiable on-surface intersection to generate on-surface points, and propose view-consistent losses on such differentiable points to regularize the neural surface learning. Based on this point, we propose a joint learning strategy, named SC-NeuS, to perform geometry-consistent surface reconstruction in an end-to-end manner. With extensive evaluation on public datasets, our SC-NeuS can achieve consistently better surface reconstruction results with fine-grained details than previous approaches, especially from sparse and noisy camera views. The source code is available at https://github.com/zouzx/sc-neus.git.

An Automatic and Robust Visual SLAM Method for Intra-Abdominal Environment Reconstruction

Three-dimensional (3D) surface reconstruction is used to solve the problem of the narrow field of view in laparoscopy. It can provide surgeons or computer-assisted surgery systems with real-time complete internal abdominal anatomy. However, rapid changes in image depth, less texture, and specular reflection pose a challenge for the reconstruction. It is difficult to stably complete the reconstruction process using feature-based simultaneous localization and mapping (SLAM) method. This paper proposes a robust laparoscopic 3D surface reconstruction method using SLAM, which can automatically select appropriate parameters for stereo matching and robustly find matching point pairs for laparoscope motion estimation. The changing trend of disparity maps is used to predict stereo matching parameters to improve the quality of the disparity map. Feature patch extraction and tracking are selected to replace feature point extraction and matching in motion estimation, which reduces its failure and interruption in feature-based SLAM. The proposed feature patch matching method is suitable for parallel computing, which can improve its computing speed. Evaluation results on public in vivo and ex vivo porcine abdominal video data show the efficiency and robustness of our 3D surface reconstruction approach.

Machine learning for phase-resolved reconstruction of nonlinear ocean wave surface elevations from sparse remote sensing data

Accurate short-term predictions of phase-resolved water wave conditions are crucial for decision-making in ocean engineering. However, the initialization of remote-sensing-based wave prediction models first requires a reconstruction of wave surfaces from sparse measurements like radar. Existing reconstruction methods either rely on computationally intensive optimization procedures or simplistic modelling assumptions that compromise the real-time capability or accuracy of the subsequent prediction process. We therefore address these issues by proposing a novel approach for phase-resolved wave surface reconstruction using neural networks based on the U-Net and Fourier neural operator (FNO) architectures. Our approach utilizes synthetic yet highly realistic training data on uniform one-dimensional grids, that is generated by the high-order spectral method for wave simulation and a geometric radar modelling approach. The investigation reveals that both models deliver accurate wave reconstruction results and show good generalization for different sea states when trained with spatio-temporal radar data containing multiple historic radar snapshots in each input. Notably, the FNO demonstrates superior performance in handling the data structure imposed by wave physics due to its global approach to learn the mapping between input and output in Fourier space.

A Bioengineering‐Regenerative Medicine Approach for Ocular Surface Reconstruction Using a Functionalized Native Cornea‐Derived Bio‐Scaffold

AbstractLimbal stem cell deficiency (LSCD) is characterized by the loss of limbal epithelial stem cells (LESCs) which compromises corneal transparency, leading to blindness. It cannot be treated with pharmacological or corneal transplantation interventions, instead a specialized stem cell (SC) therapy is needed to restore eye health and sight. Herein, a native cornea‐derived biomaterial, a by‐product of a laser refractive surgical procedure called small incision lenticule extraction is identified as a new cell delivery matrix. Culture conditions are optimized to facilitate LESC attachment, expansion and stratification, and their identity is immunophenotyped. Using electron microscopy, bio‐constructs display stratification, similar to the architectural and cellular organization of a native mammalian cornea with formation of a basement membrane and an orderly array of collagen fibrils. Neuronal growth and depleted CD45+/CD14+ leukocytes on lenticules are also shown, suggesting that in transplantation experiments, they will re‐innervate and not trigger a host‐mediated immune response. Finally, human lenticules are geometrically customized to successfully fit them over a LSCD murine cornea ex vivo, during which they maintain curvature. The authors are poised to conduced similar studies in live mice using these and other carriers currently used in the clinic to compare SC therapy outcomes.

Sat-Mesh: Learning Neural Implicit Surfaces for Multi-View Satellite Reconstruction

Automatic reconstruction of surfaces from satellite imagery is a hot topic in computer vision and photogrammetry. State-of-the-art reconstruction methods typically produce 2.5D elevation data. In contrast, we propose a one-stage method directly generating a 3D mesh model from multi-view satellite imagery. We introduce a novel Sat-Mesh approach for satellite implicit surface reconstruction: We represent the scene as a continuous signed distance function (SDF) and leverage a volume rendering framework to learn the SDF values. To address the challenges posed by lighting variations and inconsistent appearances in satellite imagery, we incorporate a latent vector in the network architecture to encode image appearances. Furthermore, we introduce a multi-view stereo constraint to enhance surface quality. This constraint minimizes the similarity between image patches to optimize the position and orientation of the SDF surface. Experimental results demonstrate that our method achieves superior visual quality and quantitative accuracy in generating mesh models. Moreover, our approach can learn seasonal variations in satellite imagery, resulting in texture mesh models with different and consistent seasonal appearances.

Learning Implicit Neural Representation for Satellite Object Mesh Reconstruction

Constructing a surface representation from the sparse point cloud of a satellite is an important task for satellite on-orbit services such as satellite docking and maintenance. In related studies on surface reconstruction from point clouds, implicit neural representations have gained popularity in learning-based 3D object reconstruction. When aiming for a satellite with a more complicated geometry and larger intra-class variance, existing implicit approaches cannot perform well. To solve the above contradictions and make effective use of implicit neural representations, we built a NASA3D dataset containing point clouds, watertight meshes, occupancy values, and corresponding points by using the 3D models on NASA’s official website. On the basis of NASA3D, we propose a novel network called GONet for a more detailed reconstruction of satellite grids. By designing an explicit-related implicit neural representation of the Grid Occupancy Field (GOF) and introducing it into GONet, we compensate for the lack of explicit supervision in existing point cloud surface reconstruction approaches. The GOF, together with the occupancy field (OF), serves as the supervised information for neural network learning. Learning the GOF strengthens GONet’s attention to the critical points of the surface extraction algorithm Marching Cubes; thus, it helps improve the reconstructed surface’s accuracy. In addition, GONet uses the same encoder and decoder as ConvONet but designs a novel Adaptive Feature Aggregation (AFA) module to achieve an adaptive fusion of planar and volume features. The insertion of AFA allows for the obtained implicit features to incorporate more geometric and volumetric information. Both visualization and quantitative experimental results demonstrate that our GONet could handle 3D satellite reconstruction work and outperform existing state-of-the-art methods by a significant margin. With a watertight mesh, our GONet achieves 5.507 CD-L1, 0.8821 F-score, and 68.86% IoU, which is equal to gains of 1.377, 0.0466, and 3.59% over the previous methods using NASA3D, respectively.

DiffSVR: Differentiable Neural Implicit Surface Rendering for Single-View Reconstruction with Highly Sparse Depth Prior

It is well-known that 3D shape and texture reconstruction from a single-view image is a very challenging and an ill-posed problem, especially without 3D supervision/ground truth. Existing neural implicit surface reconstruction approaches do easily get trapped in the local minima and cannot produce high-fidelity geometry and high-quality textures (and rendered images) under single-view setting, even with provided highly sparse depth prior. In this paper, we propose a new self-supervised learning method DiffSVR that represents a complicated surface as a new depth-aware occupancy function (DOF) and utilizes an end-to-end differentiable surface rendering paradigm to optimize the neural DOF field relying only on single-view image with highly sparse depth information. The developed surface-aware sampling, occupancy self-labeling, and differentiable surface rendering with inverse computation techniques can enhance both the neural implicit surface reconstruction and the neural renderer. The extensive experiments and comparisons on two real-world benchmark datasets (e.g., DTU and KITTI) demonstrate that our approach not only numerically outperforms the current state-of-the-art methods by a large margin, but also produces surface mesh model with qualitatively better geometric details and more accurate textures, as well as exhibits good performance on generalizability and flexibility. The code and data are available at https://github.com/akomarichev/DiffSVR.

GEMVS: a novel approach for automatic 3D reconstruction from uncalibrated multi-view Google Earth images using multi-view stereo and projective to metric 3D homography transformation

ABSTRACT This paper proposes a novel approach for automatic 3D surface reconstruction from uncalibrated and multi-view Google Earth images by using a multi-view stereo method and 3D projective to metric transformation. Without the Rational Polynomial Coefficients, it is impossible to obtain the metric reconstruction of the 3D surface from multi-view satellite images. We solve the uncalibrated multi-view satellite image problem by employing a multi-view stereo vision technique followed by a projective to metric transformation. The virtual pose parameters of the satellite images are obtained by using COLMAP, and the virtual 3D projective reconstruction is done by using EnSoft3D. For projective to metric transformation, we propose to employ 3D homography transformation. Eight 3D correspondence pairs on the viewing frustums between the virtual reference camera and the ideal nadir camera are used to derive a 3D homography matrix. Using the 3D homography matrix, we finally obtain the metric reconstruction of 3D surface up to an unknown height scale in a reference coordinate system of the Google Earth desktop software. Experiments are done in several world locations on Google Earth including building and vegetation areas. Reconstruction error analysis with the Data Fusion Contest 19 dataset is also presented. The average of MAE and RMSE of five tile regions in the dataset are 1.596 m and 2.083 m, respectively.

SurfFlow: A Flow-Based Approach for Rapid and Accurate Cortical Surface Reconstruction from Infant Brain MRI.

The infant brain undergoes rapid changes in volume, shape, and structural organization during the first postnatal year. Accurate cortical surface reconstruction (CSR) is essential for understanding rapid changes in cortical morphometry during early brain development. However, existing CSR methods, designed for adult brain MRI, fall short in reconstructing cortical surfaces from infant MRI, owing to the poor tissue contrasts, partial volume effects, and rapid changes in cortical folding patterns. Here, we introduce an infant-centric CSR method in light of these challenges. Our method, SurfFlow, utilizes three seamlessly connected deformation blocks to sequentially deform an initial template mesh to target cortical surfaces. Remarkably, our method can rapidly reconstruct a high-resolution cortical surface mesh with 360k vertices in approximately one second. Performance evaluation based on an MRI dataset of infants 0 to 12 months of age indicates that SurfFlow significantly reduces geometric errors and substantially improves mesh regularity compared with state-of-the-art deep learning approaches.

Ultrahigh Performance Self‐Supported Cl‐Cu2O@Cu Foam Electrode Fabricated by Sonochemical Surface Reconstruction Approach

AbstractUnlike powdery transition metal oxide (TMO) coated electrodes, self‐supported TMO electrodes have attracted considerable interests for optimizing the electrocatalytic performances of electrodes due to sufficient exposure of active sites and improved TMO‐substrate interfaces. However, conventional manufacture methods for self‐supported electrodes are faced with inefficiency and high complexity. Herein, a facile sonochemical surface reconstruction approach is proposed to fabricate Cl‐doped Cu2O@Cu foam (Cl‐Cu2O@Cu) electrode with ultrahigh sensitivity for glucose sensing. The intensive collapse of cavitation bubbles near Cu foam under sonication within NH4Cl solution, for this reason, massive Cu complex ions and highly active Cu surfaces are simultaneously generated. Then the nanosized Cu2O particles with large active surfaces are in situ formed and tightly anchored on the Cu surface, which undergoes Cl doping in Cu2O, resulting in ascending electrical conductivity. The Cl‐Cu2O@Cu electrode shows excellent catalytic performance to glucose, presenting ultrahigh sensitivity of 28 mA cm−2 × 10−3 m and ultralow limit of detection of 0.35 × 10−6 m (S/N = 3). This work promotes a novel approach for the rational design of self‐supported electrochemical electrodes by sonochemistry.

Labial mucosal epithelium grafting in an ex vivo human donor cornea model

The purpose of the study was to establish a simple ex vivo corneal re-epithelization model and study the labial mucosal epithelium grafting as a potential approach for ocular surface reconstruction. Four human donor corneal buttons were overstored in a corneal cold storage solution at 4 °C for 32–52 days. Four labial oral mucosa strips were dissected from four patients during fornix reconstruction after they signed informed consent. The substantia propria was trimmed off, and the resulting graft was sutured near the corneal limbus with running sutures (thus forming the tissue construct). Constructs were cultured under the standard conditions with the anterior corneal side outwards. After 3 weeks of culture, constructs were removed, washed, and fixed. Sections were stained with hematoxylin and eosin (HE), anti-keratins 4, 13, 19, and p63. Nuclei were counterstained with Hoechst. After the cultivation, all constructs were integral with the attached graft and non-loosened sutures. The native cells were absent in all donor corneas. Histological evaluation demonstrated that the labial mucosal grafts were attached to the Bowman's membrane (BM), and its cellular outgrowths were found to be transit from the graft to the BM over the anterior surface in all constructs. Cells expressed mucosal epithelial keratins 4, 13, and 19, and several were p63-positive in nuclei. In the study, a simple ex vivo corneal re-epithelization model was successfully established. The model was potent in studying the labial mucosal epithelium grafting as an option for autologous ocular surface reconstruction in patients with bilateral limbal stem cell deficiency.

Learning modified indicator functions for surface reconstruction

Surface reconstruction is a fundamental problem in 3D graphics. In this paper, we propose a learning-based approach for implicit surface reconstruction from raw point clouds without normals. Our method is inspired by Gauss Lemma in potential energy theory, which gives an explicit integral formula for the indicator functions. We design a novel deep neural network to perform surface integral and learn the modified indicator functions from un-oriented and noisy point clouds. We concatenate features with different scales for accurate point-wise contributions to the integral. Moreover, we propose a novel Surface Element Feature Extractor to learn local shape properties. Experiments show that our method generates smooth surfaces with high normal consistency from point clouds with different noise scales and achieves state-of-the-art reconstruction performance compared with current data-driven and non-data-driven approaches.

Surface Reconstruction: Roles in the Field of Computer Vision and Computer Graphics

Surface Reconstruction is the most potent aspect of 3D computer vision. It allows recapturing or imitating of the shape of real objects. It also provides sufficient knowledge regarding the mathematical foundation for rendering applications which are widely used for analyzing medical volume data, modeling, 3D interior designing, architectural designing. In our paper, we have mentioned various algorithms and approaches for surface reconstruction and their applications. Moreover, we have tried to emphasize the necessity of surface reconstruction for solving image analysis related problem.

Dynamic Three-Dimensional Surface Reconstruction Approach for Continuously Deformed Objects

For dynamic three-dimensional (3-D) surface reconstruction of continuously deformed objects, an efficient method employing one fringe pattern is proposed. First, a two-dimensional wavelet transform profilometry (2-D WTP) employing an improved cost function is adopted to obtain the wrapped phase, which has the better noise suppression ability. Then, a new 3-D phase unwrapping algorithm is introduced to obtain the unwrapped phase, which includes four steps. a) A new 3-D residues recognition method is proposed. b) A 3-D branch cut construction algorithm based on an improved GA is proposed. c) A 3-D phase derivative variance quality map is established. d) A 3-D flood filling phase unwrapping algorithm guided by 3-D quality map avoiding 3-D branch cut (3-D FGQAB) is introduced to obtain the unwrapped phase. Simulation and experimental comparisons are carried out, the results show that the proposed method can be applied to dynamic 3-D reconstruction of the continuously deformed objects.

Structural Elements Detection and Reconstruction (SEDR): A Hybrid Approach for Modeling Complex Indoor Structures

We present a hybrid approach for modeling complex interior structural elements from the unstructured point cloud without additional information. The proposed approach focuses on an integrated modeling strategy that can reconstruct structural elements and keep the balance of model completeness and quality. First, a data-driven approach detects the complete structure points of indoor scenarios including the curved wall structures and detailed structures. After applying the down-sampling process to point cloud dataset, ceiling and floor points are detected by RANSAC. The ceiling boundary points are selected as seed points of the growing algorithm to acquire points related to the wall segments. Detailed structures points are detected using the Grid-Slices analysis approach. Second, a model-driven refinement is conducted to the structure points that aims to decrease the impact of point cloud accuracy on the quality of the model. RANSAC algorithm is implemented to detect more accurate layout, and the hole in structure points is repaired in this refinement step. Lastly, the Screened Poisson surface reconstruction approach is conducted to generate the model based on the structure points after refinement. Our approach was validated on the backpack laser dataset, handheld laser dataset, and synthetic dataset, and experimental results demonstrate that our approach can preserve the curved wall structures and detailed structures in the model with high accuracy.

Corneal Recovery Following Rabbit Peripheral Blood Mononuclear Cell–Amniotic Membrane Transplantation with Antivascular Endothelial Growth Factor in Limbal Stem Cell Deficiency Rabbits

Background: Limbal stem cell deficiency (LSCD) is a refractory ocular surface disorder characterized by progressive corneal epithelial degeneration, conjunctivalization, and neovascularization, potentially leading to blindness. There are currently no effective therapeutic options for patients experiencing routine symptomatic treatment failure. Transplantation of amniotic membrane (AM) with adherent stem cells (but not bare AM transplantation alone) has shown promise in preclinical studies for ocular surface restoration. A major limitation, however, is finding a reliable stem cell source. Stem cells can be isolated from the peripheral blood mononuclear cell (PBMC) population, and these PBMC-derived stem cells have numerous advantages over allogeneic and other autologous stem cell types for therapeutic application, including relative ease of acquisition, nonimmunogenicity, and the absence of ethical issues associated with embryonic stem cells. Experiment: We examined the efficacy of autologous PBMC-AM sheet cultures combined with postoperative antiangiogenesis treatment for corneal restoration in LSCD model rabbits. Rabbit PBMCs (rPBMCs) were isolated, labeled with EdU for in vivo tracing, and then cultured on AMs in conditioned medium before transplantation. Rabbits were transplanted with bare AMs (group 1), rPBMC-AM sheets (group 2), or rPBMC-AM sheets plus postoperative treatment with the vascular endothelial growth factor antagonist bevacizumab (group 3). Corneal opacity and neovascularization were monitored by slit-lamp imaging for 8 weeks and corneas were examined histologically at 1 and 2 months. Results: Corneal opacity decreased in all three groups over 8 weeks, but was significantly lower in group 2 and even lower in group 3. Corneal neovascularization was significantly higher in group 1 throughout the observation period, and significantly lower in group 3 than group 1 and 2 by 8 weeks post-transplant. At 4 weeks, the corneal surface completed epithelialization (although thinner than normal) in group 3 but still patchy in groups 1 and 2. By 8 weeks, the epithelium in group 3 was complete and smooth, resembling a normal epithelium. Integrin β1 as a progenitor marker was also generally higher in groups 2 and 3. Conclusions: Autologous rPBMC-AM sheets with post-transplant topical bevacizumab can effectively facilitate corneal epithelium recovery in a LSCD model, suggesting clinical utility for LSCD-related ocular surface diseases. Impact statement Limbal stem cell deficiency (LSCD) increases corneal opacity and vascularization, resulting in severe visual impairment or even blindness. Traditional surgical limbal transplant is currently the main treatment option for LSCD, but carries the risks of rejection and immunosuppressant side effects. Autologous stem cell-based therapy is a promising alternative approach, but a reliable stem cell source is a major limitation. We report that transplantation of autologous rabbit peripheral blood mononuclear cell-amniotic membrane sheets plus antivascular endothelial growth factor restored avascular transparent cornea in a rabbit LSCD model. These results demonstrate a potentially effective approach for ocular surface reconstruction in bilateral LSCD.

Surface Nanostructures Fabricated by Polymerization-Induced Surface Self-Assembly

Surface properties of materials are strongly dependent on surface chemistry and surface structures. The fabrication of hierarchical surface nanostructures will endow solid surfaces with new functionalities and properties. In this research, we propose the polymerization-induced surface self-assembly (PISSA) approach for surface reconstruction. In this approach, two macro-CTAs, one grafted on silica particles and the other molecularly dissolved in solution, were used in reversible addition–fragmentation chain transfer (RAFT) dispersion polymerization, and surface micelles (s-micelles) with different morphologies and sizes on silica particles were fabricated. Kinetics studies demonstrate that there are two critical points on a plot of ln([M]0/[M]t) vs polymerization time, corresponding to the onsets of surface assembly and the self-assembly of block copolymers. The morphology of s-micelles is dependent on the monomer conversion and the length of macro-CTA. For macro-CTA with short chain length, with an increase in monomer conversion the s-micelles experience a morphology change from spherical s-micelles to layered structures. For macro-CTA with long chain length, the average size of s-micelles increases with monomer conversion. In this research, we demonstrate PISSA can be used as a versatile method for surface modification.

CuFusion2: Accurate and Denoised Volumetric 3D Object Reconstruction Using Depth Cameras

The 3D object reconstruction from depth image streams using Kinect-style depth cameras has been extensively studied. In this paper, we propose an approach for accurate camera tracking and volumetric dense surface reconstruction, assuming that a known cuboid reference object is present in the scene. Our contribution is threefold. First, we maintain the drift-free camera pose tracking by incorporating the 3D geometric constraints of the cuboid reference object into the image registration process. Second, we reformulate the problem of depth stream fusion as a binary classification problem, enabling high-fidelity surface reconstruction, especially in the concave zones of objects. Third, we further present a surface denoising strategy to mitigate the topological inconsistency (e.g., holes and dangling triangles), which facilitates the generation of a noise-free triangle mesh. We extend our public dataset CU3D with several new image sequences, test our algorithm on these sequences, and quantitatively compare them with other state-of-the-art algorithms. Both our dataset and our algorithm are available as open-source content at https://github.com/zhangxaochen/CuFusion for other researchers to reproduce and verify our results.