Cell-free massive multiple-input multiple-output (CF-mMIMO) is a promising architecture for 6G wireless networks through distributed access point (AP) cooperation. In large-scale deployments where user demand exceeds system capacity, effective user admission control (UAC) is essential to select users while meeting quality-of-service (QoS) requirements. The UAC problem in CF-mMIMO is inherently challenging, involving both discrete user selection and continuous power allocation variables. To address this challenge, we propose a Graphormer-enhanced Monte Carlo Tree Search (GE-MCTS) framework that integrates a Graphormer-based neural network (NN) with Monte Carlo Tree Search (MCTS). This framework leverages the Graphormer’s capability to model the graph-structured AP–user topology and MCTS’s planning proficiency to efficiently explore the vast decision space. Furthermore, to accommodate users initially unadmitted due to system constraints, we introduce a complementary AP deployment problem. By adapting the GE-MCTS framework, we optimize the placement of additional APs to achieve full user admission with the minimal number of new APs required. Simulation results demonstrate the effectiveness of our proposed framework. For UAC, with low computational complexity, GE-MCTS consistently admits 26.3–41.7% more users compared to baseline methods across various network scales. For AP deployment, our framework requires 45–73% fewer additional APs to achieve full user admission, highlighting its efficiency and scalability.

Clustering is a fundamental method for extracting meaningful information from large-scale genomic datasets. As sequencing technologies advance, efficient and scalable clustering tools have become increasingly important. Despite its outstanding efficiency in large-scale genome clustering tasks, RabbitTClust still faces certain limitations. On the one hand, as data volumes continue to grow, there remains room for further optimization of its computational performance. On the other hand, RabbitTClust is not well suited for frequent incremental data updates or fast clustering across multiple thresholds. To address these limitations, we introduce RabbitTClust2, a highly efficient and versatile tool designed for clustering large-scale genomic sequences. RabbitTClust2 integrates an efficient sketching algorithm, a pruningand inverted-index-based minimum spanning tree construction method, and strategies for reusing intermediate results. With these advancements, RabbitTClust2 is able to cluster the latest RefSeq bacterial dataset (195 k genomes, 820 GB in FASTA) within 5 minutes. Compared to previous versions, RabbitTClust2 achieves a$2.4 \times$to$4.5 \times$speedup while maintaining comparable clustering accuracy, with a 21 % reduction in memory consumption. On a distributed multi-node platform, RabbitTClust2 is capable of clustering 2.6 million genomes in approximately one hour. Furthermore, RabbitTClust2 offers significant versatility by supporting efficient incremental clustering and rapid multithreshold analysis. RabbitTClust2 utilizes incremental clustering to integrate 1,000 new sequences into a pre-clustered dataset of 194,000 genomes within 1 minute, a process that results in a$22.7 \times$speedup over RabbitTClust. In addition, we used RabbitTClust2 to generate a series of clustering results with Mash distance thresholds ranging from 0.01 to 0.2 (a total of 20 values) within 7 minutes on the RefSeq bacterial dataset. The results showed that when the clustering threshold approached 0.1, the cluster compositions changed significantly, suggesting that 0.1 may represent a critical threshold for genuslevel classification in bacteria. RabbitTClust2 is available at https://github.com/RabbitBio/RabbitTClust.

The demand for semantically rich 3D models of indoor scenes is rapidly growing, driven by applications in augmented reality, virtual reality, and robotics. However, creating them from sparse views remains a challenge due to geometric ambiguity. Existing methods often treat semantics as a passive feature painted on an already-formed, and potentially flawed, geometry. We posit that for robust sparse-view reconstruction, semantic understanding instead be an active, guiding force. This paper introduces AlignGS, a novel framework that actualizes this vision by pioneering a synergistic, end-to-end optimization of geometry and semantics. Our method distills rich priors from 2D foundation models and uses them to directly regularize the 3D representation through a set of novel semantic-to-geometry guidance mechanisms, including depth consistency and multifaceted normal regularization. Extensive evaluations on standard benchmarks demonstrate that our approach achieves state-of-the-art results in novel view synthesis and produces reconstructions with superior geometric accuracy. The results validate that leveraging semantic priors as a geometric regularizer leads to more coherent and complete 3D models from limited input views. Our code is avaliable at https://github.com/MediaX-SJTU/AlignGS.

Radar imaging provides non-contact, privacy-preserving, and environmentally robust monitoring for continuous human motion recognition (HMR) by leveraging diverse information embedded in various radar signal domains. However, current research has not effectively integrated multi-radar and multi-domain imaging to fully exploit the benefits of distributed radar systems. To bridge this gap, we propose a multi-radar, multi-domain parallel cross-attention model with four key components: intra-domain cross-radar weight sharing encoders specific to each domain for consistent feature extraction and parameter reduction, domain-level parallel cross-attention (DLPCAN) modules to fuse domain-specific features and enhance feature representation robustness in each radar, a source-level attention fusion (SLAF) module to highlight significant features from multiple radar inputs, and two bi-directional gated recurrent unit (BiGRU) modules to capture temporal information. The model is trained using connectionist temporal classification (CTC) loss for effective sequence prediction. By integrating data from multiple radar nodes and domains, our approach significantly improves continuous HMR performance compared to single radar systems and single domain data. Comparative evaluations demonstrate that our model outperforms state-of-the-art radar imaging-based HMR solutions.

Novel-view synthesis with sparse input views is important for real-world applications like AR/VR and autonomous driving. Recent methods have integrated depth information into NeRFs for sparse input synthesis, leveraging depth prior for geometric and spatial understanding. However, most existing works tend to overlook inaccuracies within depth maps and have low time efficiency. To address these issues, we propose a depth-guided robust and fast point cloud fusion NeRF for sparse inputs. We perceive radiance fields as an explicit voxel grid of features. A point cloud is constructed for each input view, characterized within the voxel grid using matrices and vectors. We accumulate the point cloud of each input view to construct the fused point cloud of the entire scene. Each voxel determines its density and appearance by referring to the point cloud of the entire scene. Through point cloud fusion and voxel grid fine-tuning, inaccuracies in depth values are refined or substituted by those from other views. Moreover, our method can achieve faster reconstruction and greater compactness through effective vector-matrix decomposition. Experimental results underline the superior performance and time efficiency of our approach compared to state-of-the-art baselines.

Recently, NVS in human-object interaction scenes has received increasing attention. Existing human-object interaction datasets mainly consist of static data with limited views, offering only RGB images or videos, mostly containing interactions between a single person and objects. Moreover, these datasets exhibit complexities in lighting environments, poor synchronization, and low resolution, hindering high-quality human-object interaction studies. In this paper, we introduce a new people-object interaction dataset that comprises 38 series of 30-view multi-person or single-person RGB-D video sequences, accompanied by camera parameters, foreground masks, SMPL models, some point clouds, and mesh files. Video sequences are captured by 30 Kinect Azures, uniformly surrounding the scene, each in 4 K resolution 25 FPS, and lasting for 1~19 seconds. Meanwhile, we evaluate some SOTA NVS models on our dataset to establish the NVS benchmarks. We hope our work can inspire further research in human-object interaction.

Novel-view synthesis with sparse input views is important for practical applications such as AR/VR and autonomous driving. Many works in this field have already integrated depth information into NeRF, utilizing depth priors for assistance in geometric and spatial understanding. However, most existing work tends to either overlook the inaccuracies in depth maps or only handle them roughly, limiting the effectiveness of the synthesis. To address this issue, we propose a depth-guided robust point cloud fusion NeRF for sparse input synthesis. We first construct a point cloud for each input view, with a novel point cloud representation based on learnable matrices and vectors. Then, through an additional lightweight scene fusion network, we fuse the point clouds from each input view to build a point cloud of the entire scene. By optimizing the point cloud representation and scene fusion network, inaccuracies in the depth map can be adjusted and refined, thereby achieving a more precise perception of the overall scene. Each voxel in the scene is determined by referencing the fused point cloud to establish its density and appearance. Experimental results demonstrate that our method outperforms state-of-the-art baselines.

Crowd counting, which has been widely adopted for estimating the number of people in safety-critical scenes, is shown to be vulnerable to adversarial examples in the physical world (e.g., adversarial patches). Though harmful, adversarial examples are also valuable for evaluating and better understanding model robustness. However, existing adversarial example generation methods for crowd counting lack strong transferability among different black-box models, which limits their practicability for real-world systems. Motivated by the fact that attacking transferability is positively correlated to the model-invariant characteristics, this paper proposes the Perceptual Adversarial Patch (PAP) generation framework to tailor the adversarial perturbations for crowd counting scenes using the model-shared perceptual features. Specifically, we handcraft an adaptive crowd density weighting approach to capture the invariant scale perception features across various models and utilize the density guided attention to capture the model-shared position perception. Both of them are demonstrated to improve the attacking transferability of our adversarial patches. Extensive experiments show that our PAP could achieve state-of-the-art attacking performance in both the digital and physical world, and outperform previous proposals by large margins (at most +685.7 MAE and +699.5 MSE). Besides, we empirically demonstrate that adversarial training with our PAP can benefit the performance of vanilla models in alleviating several practical challenges in crowd counting scenarios, including generalization across datasets (up to -376.0 MAE and -354.9 MSE) and robustness towards complex backgrounds (up to -10.3 MAE and -16.4 MSE).