Despite significant progress in point cloud upsampling, most existing methods rely heavily on supervised training with paired data, which are often difficult to acquire. Moreover, the inherent lack of explicit connectivity in point clouds makes it difficult to achieve both continuous and uniform densification while accurately recovering fine geometric structures. To address these problems, we propose an upsampling model that treats this task as saliency-aware implicit surface sampling, enabling self-supervised and fine-grained point densification. Central to our idea is correlating implicit surface reconstruction with salient point identification, and carrying out sampling on the saliency-aware surface representation. Motivated by this, we introduce a salient point detector along with a corresponding implicit surface-based interpolator, and a geometry filter, upon which we develop a three-stage architecture consisting of a pre-trained saliency guidance block, a saliency-aware enhancer, and an upsampler. The guidance block captures meaningful shape patterns to prevent detail loss and incomplete recovery, while the enhancer facilitates detail enhancement in complex salient regions. These components are integrated with the upsampler to generate dense results that accurately retain both global shape and meticulous structures. In comparison with state-of-the-art methods, our model significantly improves the upsampling quality. Extensive experiments conducted on various datasets, comprising both synthetic and real-world captured shapes, demonstrate the flexibility and availability of our method in processing the point clouds across different scales and distributions.

Utilizing visualization techniques for opinion mining and sentiment exploration of social media texts is a pivotal concern in the visualization field. Research predominantly concentrates on the collective expression of numerous users on popular topics within the public domain. However, there is a need to focus on changing or evolving sentiments for individuals. This paper tries to integrate sentiment analysis with visualization techniques to enable individual users to visually explore the underlying sentiment stories behind their social media texts. Therefore, the SentiSand system is designed and developed based on user requirements. It employs innovative visual design to transform the text from users' social media into a series of visual stories to facilitate users tracking their sentiment changes. The system provides the functionalities of visualizing the overview of the sentiment evolution over a specified period, the sentiment polarity distribution, and words used over that period as well. In a user study, eighteen participants with diverse backgrounds are invited to evaluate the system. The results demonstrate that SentiSand can effectively and intuitively depict sentiment states and changes. All participants gave the system high ratings across dimensions of creativity, practicality, and narrative coherence. Overall, this paper introduces a novel perspective and methodology for visualizing personal sentiments, thereby contributing to an enhanced comprehension of and intervention in individual sentiment states.

In the digital era, traditional methods of historical learning are undergoing a transition, as there is a growing demand for intuitive and visually engaging approaches. Vi-sualization, particularly timeline-based visualization, is critical in bridging the gap between extensive and complex historical events and history learners due to its intuitive features. Despite the advances in existing timeline visualization tools, they often face challenges: visualization creators spend much time in the materials preparation stage, especially for designers unfamiliar with the history; the lack of flexible interaction limits users from designing customized representations, which decreases the user's interest. To address these issues, we have developed Create3DHistory, a system that leverages the advanced Artificial Intelligence Generated Content and visualization technologies, enabling users to easily generate historical events content and create 3D and 2D historical event timeline representation. Users only need to input the name of a historical event, and the system will automatically generate corresponding content and deploy it onto the timeline representation. Moreover, users can edit the generated content or upload self-prepared multimedia materials for customized event visualization. The user study has demonstrated that Create3DHistory can effectively assist users in customizing personalized historical event timelines. Overall, the proposed tool simplifies the timeline creation process using AIGC technology, providing an overview and detailed information on-demand functions for various tasks.

The digitization of cultural heritage offers opportunities for preservation and dissemination but also presents challenges in fostering public engagement and narrative coherence. This paper proposes an interactive storytelling system that integrates AI technologies, including AI agents, LLMs, and visualization tools, to enhance user engagement with cultural heritage. Using the Jataka story murals from the Dunhuang grottoes as a case study, the proposed system features three core modules: mural segmentation with cultural context explanations, an interactive knowledge graph that enables users to engage with the murals through storytelling, and user-driven creative reconstruction through comic-style narratives. An AI agent supports these modules by offering conversational interactions that provide real-time explanations, personalized guidance, and creative suggestions. User evaluations based on the User Experience Questionnaire demonstrate the system’s effectiveness in promoting user engagement, with particularly high scores in attractiveness, novelty, and stimulation. These results suggest that the system increases accessibility and interactivity, helping users form deeper emotional and intellectual connections with cultural heritage. The prototype system demonstrates the framework’s potential to make cultural heritage more accessible and participatory, fostering richer connections through interaction, storytelling, and creative expression. These findings highlight the potential of this approach to connect traditional heritage with contemporary audiences, offering new avenues for participation, learning, and creative engagement.

Learning 3-D structures from incomplete point clouds with extreme sparsity and random distributions is a challenge since it is difficult to infer topological connectivity and structural details from fragmentary representations. Missing large portions of informative structures further aggravates this problem. To overcome this, a novel graph convolutional network (GCN) called dynamic and structure-aware NETwork (DSANet) is presented in this article. This framework is formulated based on a pyramidic auto-encoder (AE) architecture to address accurate structure reconstruction on the sparse and incomplete point clouds. A PointNet-like neural network is applied as the encoder to efficiently aggregate the global representations of coarse point clouds. On the decoder side, we design a dynamic graph learning module with a structure-aware attention (SAA) to take advantage of the topology relationships maintained in the dynamic latent graph. Relying on gradually unfolding the extracted representation into a sequence of graphs, DSANet is able to reconstruct complicated point clouds with rich and descriptive details. To associate analogous structure awareness with semantic estimation, we further propose a mechanism, called structure similarity assessment (SSA). This method allows our model to surmise semantic homogeneity in an unsupervised manner. Finally, we optimize the proposed model by minimizing a new distortion-aware objective end-to-end. Extensive qualitative and quantitative experiments demonstrate the impressive performance of our model in reconstructing unbroken 3-D shapes from deficient point clouds and preserving semantic relationships among different regional structures.

In bloodstain examination of collaborative medicine and forensics, the analysis and identification of time since deposition (TSD) plays a significant role. Traditional bloodstain analysis methods can only provide a rough estimate for the TSD of traces, and they are time-consuming. To address this issue, we propose a lightweight framework called Fourier Transform Infrared Network (FTIR-Net) that combines wavelet transform with deep learning. To be specific, we parallelly perform wavelet transform on infrared spectra and compute its second derivative to attain the sequential signal and spectral image. Then, the learning component employs two separate branches to extract features from the one-dimensional (1D) spectra signal and two-dimensional (2D) coefficient images provided by continuous wavelet transform (CWT). To effectively aggregate information from the spectral image, we design a Squeeze-and-Excitation Network (SENet) and combine it with 2D convolution. Finally, the extracted features are concatenated and flattened, followed by two fully connected (FC) layers for retention time analysis. Since the standard bloodstain dataset is lacking, we create a dataset that associates bloodstain with the attenuated total reflectance of Fourier transform infrared (ATR-FTIR). To demonstrate the effectiveness of our model in bloodstain analysis and exploit the properties of the proposed dataset, we present comprehensive experiments and ablation studies.

A large number of collaborative manufacturing tasks are directly performed on point clouds. With the growing size of point clouds, the computational demands of these tasks also increase. One possible solution is to sample the point clouds. The most commonly used sampling method is farthest point sampling, but it does not consider downstream tasks, often leading to sampling non-informative points for the tasks. With the development of neural networks, various methods have been proposed to sample point clouds in a task-oriented learning manner. However, most methods are based on generation rather than selecting a subset of point clouds. In this work, we propose a novel adaptive keypoint sampling method, called MGE-Net, that combines neural network-based learning with direct point selection based on multi-scale geometry estimation. In addition, we design a feature extraction module based on multi-scale attention graph convolution to provide accurate information for subsequent keypoint detection. Relying on the contribution of point clouds to the task, our framework aims to sample a subset of point clouds specifically optimized for downstream tasks. Both qualitative and quantitative experimental results demonstrate that our sampling method exhibits superior performance in common point cloud classification and segmentation tasks.

Despite the success of recent upsampling approaches, generating high-resolution point sets with uniform distribution and meticulous structures is still challenging. Unlike existing methods that only take spatial information of the raw data into account, we regard point cloud upsampling as generating dense point clouds from deformable topology. Motivated by this, we present SPU-PMD, a self-supervised topological mesh deformation network, for 3D densification. As a cascaded framework, our architecture is formu-lated by a series of coarse mesh interpolator and mesh de-formers. At each stage, the mesh interpolator first produces the initial dense point clouds via mesh interpolation, which allows the model to perceive the primitive topology better. Meanwhile, the deformer infers the morphing by estimating the movements of mesh nodes and reconstructs the de-scriptive topology structure. By associating mesh deformation with feature expansion, this module progressively re-fines point clouds' surface uniformity and structural details. To demonstrate the effectiveness of the proposed method, extensive quantitative and qualitative experiments are con-ducted on synthetic and real-scanned 3D data. Also, we compare it with state-of-the-art techniques to further illus-trate the superiority of our network. The project page is: https://github.com/lyz21/spU-PMd.

Point clouds are often sparse and incomplete in real-world scenarios. The prevailing methods for point cloud completion typically rely on encoding the partial points and then decoding complete points from a global feature vector, which might lose the existing patterns and elaborate structures. To address these issues, we propose WalkFormer, a novel approach to predict complete point clouds through a partial deformation process. Concretely, our method samples locally dominant points based on feature similarity and moves the points to form the missing part. Since these points maintain representative information of the surrounding structures, they are appropriately selected as the starting points for multiple guided walks. Furthermore, we design a Route Transformer module to exploit and aggregate the walk information with topological relations. These guided walks facilitate the learning of long-range dependencies for predicting shape deformation. Qualitative and quantitative evaluations demonstrate that our proposed approach achieves superior performance compared to state-of-the-art methods in the 3D point cloud completion task.

Point clouds are fundamental in the representation of 3D objects. However, they can also be highly unstructured and irregular. This makes it difficult to directly extend 2D generative models to three-dimensional space. In this article, we cast the problem of point cloud generation as a topological representation learning problem. In order to capture the representative features of 3D shapes in the latent space, we propose a hierarchical mixture model that integrates self-attention with an inference tree structure for constructing a point cloud generator. Based on this, we design a novel Generative Adversarial Network (GAN) architecture that is capable of generating recognizable point clouds in an unsupervised manner. The proposed adversarial framework (SG-GAN) relies on self-attention mechanism and Graph Convolution Network (GCN) to hierarchically infer the latent topology of 3D shapes. Embedding and transferring the global topology information in a tree framework allows our model to capture and enhance the structural connectivity. Furthermore, the proposed architecture endows our model with partially generating 3D structures. Finally, we propose two gradient penalty methods to stabilize the training of SG-GAN and overcome the possible mode collapse of GAN networks. To demonstrate the performance of our model, we present both quantitative and qualitative evaluations and show that SG-GAN is more efficient in training and it exceeds the state-of-the-art in 3D point cloud generation.

Point clouds are the most general data representations of real and abstract objects, and have a wide variety of applications in many science and engineering fields. Point clouds also provide the most scalable multi-resolution composition for geometric structures. Although point cloud learning has shown remarkable results in shape estimation and semantic segmentation, the unsupervised generation of 3D object parts still pose significant challenges in the 3D shape understanding problem. We address this problem by proposing a novel Generative Adversarial Network (GAN), named HSGAN, or Hierarchical Self-Attention GAN, with remarkable properties for 3D shape generation. Our generative model takes a random code and hierarchically transforms it into a representation graph by incorporating both Graph Convolution Network (GCN) and self-attention. With embedding the global graph topology in shape generation, the proposed model takes advantage of the latent topological information to fully construct the geometry of 3D object shapes. Different from the existing generative pipelines, our deep learning architecture articulates three significant properties HSGAN effectively deploys the compact latent topology information as a graph representation in the generative learning process and generates realistic point clouds, HSGAN avoids multiple discriminator updates per generator update, and HSGAN preserves the most dominant geometric structures of 3D shapes in the same hierarchical sampling process. We demonstrate the performance of our new approach with both quantitative and qualitative evaluations. We further present a new adversarial loss to maintain the training stability and overcome the potential mode collapse of traditional GANs. Finally, we explore the use of HSGAN as a plug-and-play decoder in the auto-encoding architecture.