Gaussian splatting has gained attention for its efficient representation and rendering of 3D scenes using continuous Gaussian primitives. However, it struggles with sparse-view inputs due to limited geometric and photometric information, causing ambiguities in depth, shape, and texture. We propose GBR: Generative Bundle Refinement, a method for high-fidelity Gaussian splatting and meshing using only 4–6 input views. GBR integrates a neural bundle adjustment module to enhance geometry accuracy and a generative depth refinement module to improve geometry fidelity. More specifically, the neural bundle adjustment module integrates a foundation network to produce initial 3D point maps and point matches from unposed images, followed by bundle adjustment optimization to improve multiview consistency and point cloud accuracy. The generative depth refinement module employs a diffusion-based strategy to enhance geometric details and fidelity while preserving the scale. Finally, for Gaussian splatting optimization, we propose a multimodal loss function incorporating depth and normal consistency, geometric regularization, and pseudo-view supervision, providing robust guidance under sparse-view conditions. Experiments on widely used datasets show that GBR significantly outperforms existing methods under sparse-view inputs. Additionally, GBR demonstrates the ability to reconstruct and render large-scale real-world scenes, such as the Pavilion of Prince Teng and the Great Wall, with remarkable details using only 6 views. More results can be found on our project page https://gbrnvs.github.io.

Website fingerprinting attacks leverage encrypted traffic features to identify specific services accessed by users within anonymity networks such as Tor. Although existing WF methods achieve high accuracy on static datasets using deep learning techniques, they struggle in dynamic environments where anonymous websites continually evolve. These methods typically require full retraining on composite datasets, resulting in substantial computational and storage burdens, and are particularly vulnerable to classification bias caused by data imbalance and concept drift. To address these challenges, we propose EIL-WF, a dynamic WF framework based on incremental learning that enables efficient adaptation to newly emerging websites without the need for full retraining. EIL-WF incrementally trains lightweight, independent classifiers for new website classes and integrates them through classifier normalization and energy alignment strategies grounded in energy-based model theory, thereby constructing a unified and robust classification model. Comprehensive experiments on two public Tor traffic datasets demonstrate that EIL-WF outperforms existing incremental learning methods by 6.2%–20.2% in identifying new websites and reduces catastrophic forgetting by 5.4%–20%. Notably, EIL-WF exhibits strong resilience against data imbalance and concept drift, maintaining stable classification performance across evolving distributions. Furthermore,EIL-WF decreases training time during model updates by 2–3 orders of magnitude, demonstrating substantial advantages over conventional full retraining paradigms.

Metalenses offer significant potential for ultra-compact computational imaging but face challenges from complex optical degradation and computational restoration difficulties. Existing methods typically rely on precise optical calibration or massive paired datasets, which are non-trivial for real-world imaging systems. Furthermore, a lack of control over the inference process often results in undesirable hallucinated artifacts. We introduce Degradation-Modeled Multipath Diffusion for tunable metalens photography, leveraging powerful natural image priors from pretrained models instead of large datasets. Our framework uses positive, neutral, and negative-prompt paths to balance high-frequency detail generation, structural fidelity, and suppression of metalens-specific degradation, alongside \textit{pseudo} data augmentation. A tunable decoder enables controlled trade-offs between fidelity and perceptual quality. Additionally, a spatially varying degradation-aware attention (SVDA) module adaptively models complex optical and sensor-induced degradation. Finally, we design and build a millimeter-scale MetaCamera for real-world validation. Extensive results show that our approach outperforms state-of-the-art methods, achieving high-fidelity and sharp image reconstruction. More materials: https://dmdiff.github.io/.

Compressive spectral image reconstruction is a critical method for acquiring images with high spatial and spectral resolution. Current advanced methods, which involve de-signing deeper networks or adding more self-attention mod-ules, are limited by the scope of attention modules and the irrelevance of attentions across different dimensions. This leads to difficulties in capturing non-local mutation features in the spatial-spectral domain and results in a signif-icant parameter increase but only limited performance im-provement. To address these issues, we propose SPECAT, a SPatial-spEctral Cumulative-Attention Transformer de-signed for high-resolution hyperspectral image reconstruction. SPECAT utilizes Cumulative-Attention Blocks (CABs) within an efficient hierarchical framework to extract features from non-local spatial-spectral details. Furthermore, it employs a projection-object Dual-domain Loss Function (DLF) to integrate the optical path constraint, a physical aspect often overlooked in current methodologies. Ulti-mately, SPECAT not only significantly enhances the reconstruction quality of spectral details but also breaks through the bottleneck of mutual restriction between the cost and accuracy in existing algorithms. Our experimental re-sults demonstrate the superiority of SPECAT, achieving 40.3 dB in hyperspectral reconstruction benchmarks, out-performing the state-of-the-art (SOTA) algorithms by 1.2 dB while using only 5% of the network parameters and 10% of the computational cost. The code is available at https://github.com/THU-luvisionISPECAT.

Mobile IP geolocation aims to obtain a mobile device’s geographic location by IP. This technology is widely used in preventing financial risk, investigating cybercrime, and delivering targeted information. Currently, there are three types of IP geolocation: based on cooperation, querying in database, or network measurement. However, since restricted cooperation, low-reliability databases, and unresponsive mobile IPs, existing technologies are hard to geolocate fine-grained location of mobile IP. In this paper, we propose the concept of district anchor, and propose a non-cooperative mobile IP geolocation scheme, including three parts: acquiring district anchors by clustering, evaluating the reliability of district anchors, and geolocating mobile IPs. We also give implemented approach of this scheme. Instead of using existing clustering algorithms treating IPs and geolocations in no particular order, we propose two-stages clustering algorithm (IPG2C) to acquire district anchors, and establish reliability evaluation mechanism by IP distribution and spatial distribution of cluster. Eventually, using obtained reliable district anchors, we use “subnet geolocation” strategy to geolocate mobile IPs. The experimental results in 10 cities show that: 1) our scheme can be used to geolocate mobile IPs without cooperation; 2) the mean geolocation error is 12.47km, where precision of 56.67% of mobile IPs is street-level and minimum error is only 13m; 3) that the mean geolocation error of the anchor-based method is smaller than that of the landmark-based method; 4) compared with 13 clustering algorithms (e.g., K-Means++, Mean Shift, DBSCAN, and GMM), mean geolocation error using IPG2C’s district anchors is reduced by 26.62%~50.77%.

Multiview stereopsis (MVS) methods, which can reconstruct both the 3D geometry and texture from multiple images, have been rapidly developed and extensively investigated from the feature engineering methods to the data-driven ones. However, there is no dataset containing both the 3D geometry of large-scale scenes and high-resolution observations of small details to benchmark the algorithms. To this end, we present GigaMVS, the first gigapixel-image-based 3D reconstruction benchmark for ultra-large-scale scenes. The gigapixel images, with both wide field-of-view and high-resolution details, can clearly observe both thePalace-scale scene structure andRelievo-scale local details. The ground-truth geometry is captured by the laser scanner, which covers ultra-large-scale scenes with an average area of 8667 m$^2$and a maximum area of 32007 m$^2$. Owing to the extremely large scale, complex occlusion, and gigapixel-level images, GigaMVS exposes problems that emerge from the poor scalability and efficiency of the existing MVS algorithms. We thoroughly investigate the state-of-the-art methods in terms of geometric and textural measurements, which point to the weakness of the existing methods and promising opportunities for future works. We believe that GigaMVS can benefit the community of 3D reconstruction and support the development of novel algorithms balancing robustness, scalability and accuracy.

The ability of camera arrays to efficiently capture higher space-bandwidth product than single cameras has led to various multiscale and hybrid systems. These systems play vital roles in computational photography, including light field imaging, 360 VR camera, gigapixel videography, etc. One of the critical tasks in multiscale hybrid imaging is matching and fusing cross-resolution images from different cameras under perspective parallax. In this paper, we investigate the reference-based super-resolution (RefSR) problem associated with dual-camera or multi-camera systems. RefSR consists of super-resolving a low-resolution (LR) image given an external high-resolution (HR) reference image, where they suffer both a significant resolution gap ( 8×) and large parallax ( ∼ 10% pixel displacement). We present CrossNet++, an end-to-end network containing novel two-stage cross-scale warping modules, image encoder and fusion decoder. The stage I learns to narrow down the parallax distinctively with the strong guidance of landmarks and intensity distribution consensus. Then the stage II operates more fine-grained alignment and aggregation in feature domain to synthesize the final super-resolved image. To further address the large parallax, new hybrid loss functions comprising warping loss, landmark loss and super-resolution loss are proposed to regularize training and enable better convergence. CrossNet++ significantly outperforms the state-of-art on light field datasets as well as real dual-camera data. We further demonstrate the generalization of our framework by transferring it to video super-resolution and video denoising.

We present PANDA, the first gigaPixel-level humAN-centric viDeo dAtaset, for large-scale, long-term, and multi-object visual analysis. The videos in PANDA were captured by a gigapixel camera and cover real-world scenes with both wide field-of-view (~1 square kilometer area) and high-resolution details (~gigapixel-level/frame). The scenes may contain 4k head counts with over 100× scale variation. PANDA provides enriched and hierarchical ground-truth annotations, including 15,974.6k bounding boxes, 111.8k fine-grained attribute labels, 12.7k trajectories, 2.2k groups and 2.9k interactions. We benchmark the human detection and tracking tasks. Due to the vast variance of pedestrian pose, scale, occlusion and trajectory, existing approaches are challenged by both accuracy and efficiency. Given the uniqueness of PANDA with both wide FoV and high resolution, a new task of interaction-aware group detection is introduced. We design a `global-to-local zoom-in' framework, where global trajectories and local interactions are simultaneously encoded, yielding promising results. We believe PANDA will contribute to the community of artificial intelligence and praxeology by understanding human behaviors and interactions in large-scale real-world scenes. PANDA Website: http://www.panda-dataset.com.

Creating virtual reality (VR) content with effective imaging systems has attracted significant attention worldwide following the broad applications of VR in various fields, including entertainment, surveillance, sports, etc. However, due to the inherent trade-off between field-of-view and resolution of the imaging system as well as the prohibitive computational cost, live capturing and generating multiscale 360° 3D video content at an eye-limited resolution to provide immersive VR experiences confront significant challenges. In this work, we propose Multiscale-VR, a multiscale unstructured camera array computational imaging system for high-quality gigapixel 3D panoramic videography that creates the six-degree-of-freedom multiscale interactive VR content. The Multiscale-VR imaging system comprises scalable cylindrical-distributed global and local cameras, where global stereo cameras are stitched to cover 360° field-of-view, and unstructured local monocular cameras are adapted to the global camera for flexible high-resolution video streaming arrangement. We demonstrate that a high-quality gigapixel depth video can be faithfully reconstructed by our deep neural network-based algorithm pipeline where the global depth via stereo matching and the local depth via high-resolution RGB-guided refinement are associated. To generate the immersive 3D VR content, we present a three-layer rendering framework that includes an original layer for scene rendering, a diffusion layer for handling occlusion regions, and a dynamic layer for efficient dynamic foreground rendering. Our multiscale reconstruction architecture enables the proposed prototype system for rendering highly effective 3D, 360° gigapixel live VR video at 30 fps from the captured high-throughput multiscale video sequences. The proposed multiscale interactive VR content generation approach by using a heterogeneous camera system design, in contrast to the existing single-scale VR imaging systems with structured homogeneous cameras, will open up new avenues of research in VR and provide an unprecedented immersive experience benefiting various novel applications.

Presenting high-resolution (HR) human appearance is always critical for the human-centric videos. However, current imagery equipment can hardly capture HR details all the time. Existing super-resolution algorithms barely mitigate the problem by only considering universal and low-level priors of image patches. In contrast, our algorithm is under bias towards the human body super-resolution by taking advantage of high-level prior defined by HR human appearance. Firstly, a motion analysis module extracts inherent motion pattern from the HR reference video to refine the pose estimation of the low-resolution (LR) sequence. Furthermore, a human body reconstruction module maps the HR texture in the reference frames onto a 3D mesh model. Consequently, the input LR videos get super-resolved HR human sequences are generated conditioned on the original LR videos as well as few HR reference frames. Experiments on an existing dataset and real-world data captured by hybrid cameras show that our approach generates superior visual quality of human body compared with the traditional method.

We present a multi-scale camera array to capture and synthesize gigapixel videos in an efficient way. Our acquisition setup contains a reference camera with a short-focus lens to get a large field-of-view video and a number of unstructured long-focus cameras to capture local-view details. Based on this new design, we propose an iterative feature matching and image warping method to independently warp each local-view video to the reference video. The key feature of the proposed algorithm is its robustness to and high accuracy for the huge resolution gap (more than 8x resolution gap between the reference and the local-view videos), camera parallaxes, complex scene appearances and color inconsistency among cameras. Experimental results show that the proposed multi-scale camera array and cross resolution video warping scheme is capable of generating seamless gigapixel video without the need of camera calibration and large overlapping area constraints between the local-view cameras.

This paper proposes a virtual 3D eyeglasses try-on system driven by a 2D Internet image of a human face wearing with a pair of eyeglasses. The main technical challenge of this system is the automatic 3D eyeglasses model reconstruction from the 2D glasses on a frontal human face. Against this challenge, this paper first proposes an eyeglasses segmentation method using a convolutional neural network-based parsing algorithm to label the glasses pixels, followed with a proposed symmetry-based level-set optimization algorithm to refine the contour of the eyeglasses. With the precisely extracted silhouette image, we take advantages of the smoothness and the symmetry priors of the eyeglasses, and propose a silhouette-based 3D deformation method to deform a 3D eyeglasses model selected from a predefined 3D model database. The obtained model is plausibly approximated to the input 2D eyeglasses after a texture mapping step. Finally, we develop a virtual try-on system to interactively synthesize the reconstructed eyeglasses on a moving target face in real time. The experimental results demonstrate the efficient and convincing virtual try-on performance of our approach and the commercial potential of our proposed system.

This paper presents some results on isogeometric analysis (IGA) using a new class of basis functions, named truncated interpolatory basis function (TIBF). While much work in IGA has been done based on approximating basis functions, such as B-splines, NURBS and T-splines, little work is reported so far using interpolatory schemes. In this article, a truncated interpolatory basis function is proposed for isogeometric analysis. The basis function has many nice properties which are desirable for both modeling and isogeometric analysis. It has also the potential to be used as an alternative basis for interpolatory subdivision schemes whose basis functions are often not explicitly available. Based on TIBF, a cut-off rule is also proposed for handling shapes with boundaries. The proposed interpolatory scheme based on TIBF exhibits modeling qualities that are pretty close to the Kobbelt subdivision scheme for a regular control mesh. Several examples on IGA based on TIBF are also presented in this article. Both the modeling and simulation examples show that TIBF provides a good alternative option for isogeometric analysis.