Siqi Li

In recent years, point regression-based crowd counting has demonstrated remarkable advantages for accurate counting in highly congested scenes. However, existing single-pass regression frameworks lack an explicit self-correction mechanism, making it difficult to rectify localization shifts or even false positives induced by local visual ambiguity in heavily occluded regions. Moreover, straightforward regression predictions struggle to capture complex cluster-level higher-order relations in dense crowds, limiting the model’s ability to leverage global context to resolve semantic confusion. To address these issues, we reformulate point regression as an iterative optimization process and propose a Hypergraph-induced Iterative Reasoning Network, HIRNet. Specifically, we first design a 2D bidirectional hypergraph iterative reasoning module (HIR-2D), which treats intermediate-round regression points as an explicit structural prior and dynamically constructs bidirectional hypergraphs along the spatial and channel dimensions. By performing higher-order feature aggregation in both dimensions, HIR-2D enforces long-range spatial consistency and recalibrates channel features, thereby effectively suppressing background distractions. Second, to mitigate the matching jitter of regressed point sets across iterations, we introduce an Anchor-Matching Iterative Deep Supervision strategy that locks the bipartite matching from the final round to ensure consistent gradient directions across rounds. Furthermore, we devise a Monotonic Improvement Loss that explicitly constrains the model to progressively reduce geometric errors throughout iterations, stabilizing the reasoning dynamics. Extensive experiments on three mainstream datasets demonstrate that our method substantially outperforms current advanced approaches.

Event-based monocular depth estimation is crucial for applications such as autonomous driving, obstacle avoidance, and navigation under high-speed scenarios. Events exhibit a unique and irregular modality. To adapt them to neural networks, some studies convert event streams into event voxels or other frame-like representations. However, these approaches tend to lose the temporal characteristics of events. In this study, we propose a network that aggregates global voxel and per-channel temporal local features of event voxels across the temporal dimension, explicitly extracting events’ temporal information. Furthermore, as noise in events can interfere with the training process and is more difficult to predict than that in images, we utilize the uncertainty estimation module to mitigate the impact of uncertain factors and enhance the robustness of the model. Additionally, we employ multi-level depth features for supervisory training, which improves prediction performance compared to methods relying solely on ground-truth depth supervision. Experiments on open source datasets demonstrate the effectiveness of the proposed method. Our code can be found at https://github.com/WuShangjie/GLUNET .

Point cloud registration (PCR) has been an important research subject for many years but remains an open problem, presenting numerous challenges. The stability of existing registration methods is often inadequate, particularly in scenarios with low overlap. This issue primarily arises from the insufficient distinctiveness of extracted point cloud features, leading to ambiguous matches and the proliferation of outliers. To address these bottlenecks in point cloud registration, it is crucial to fully leverage the color information of the point clouds to discern point correspondences effectively. However, excessive reliance on color information during the registration process may impede the recognition of point cloud geometric structures, making it essential to find a balance between the aggressiveness and stability of color integration. To tackle these challenges, we propose ColorPCR++, which unlocks the potential of color information while maintaining stability. Specifically, we design a Curvature-Color Fusion Module (CCF) to initially introduce color, which leverages curvature to enhance geometric features and balance color stability. Additionally, to further incorporate color into scene structure representation, we model the point-wise hue gradient field (HGF) for positional embedding, which proves especially effective in lowoverlap scenarios. CCF and HGF fully harness the power of color, enabling the identification of most corresponding point pairs with prominent geometric and color features. However, the reliance on color can also introduce false correspondences, reducing the inlier ratio within the correspondences obtained through feature matching. Therefore, we assess the higher-order compatibility of correspondences using Feature-based Compatibility Hypergraph Convolution (FCH), effectively filtering out outliers and accurately estimating the transformation. Evaluation across multiple datasets has demonstrated the state-of-the-art performance of ColorPCR++.

Aerial skiing is a challenging human-centric sport characterized by rapid motion, large-scale variations, and frequent occlusions. Its extensive spatial range is typically captured by cameras or drones from multiple perspectives, resulting in frequent and complex viewpoint shifts. These challenges encompass nearly all difficulties inherent in human-centric tracking tasks. In this article, we introduce SkiTrack , the first dataset explicitly designed for tracking in aerial skiing. SkiTrack enhances the performance of existing tracking algorithms across a range of human-centric scenarios by providing precise annotations. We observe distinct characteristics in the tracked components, with the skis being rigid and low in visibility and the athlete’s body highly deformable but more visible. To leverage these differences, we propose a components decoupled loss that applies separate constraints to the tracking of the athlete and skis, thereby improving tracking accuracy in skiing scenes. Our experimental results validate the effectiveness of both the SkiTrack dataset and the proposed decoupled loss function, demonstrating consistent improvements in the performance of established models on human-centric tracking tasks. Data are available at https://github.com/xiaozhangfangyang/FineSkiing .

Semantic labels are inherently tied to geometry and luminance reconstruction, as entities with similar shapes and appearances often share categories. Traditional methods use synthesis-analysis, NeRF, or 3D Gaussian representations to encode semantics and geometry separately. However, 2D methods lack view consistency, NeRF extensions are slow, and faster 3D Gaussian methods risk spatial and channel inconsistencies between semantic and RGB. Moreover, these methods require costly manual dense semantic labels. To alleviate resource demands and achieve effective semantic reconstruction with sparse inputs while enhancing RGB rendering quality, we build upon 3D Gaussian by integrating semantic features from pre-trained models-requiring no additional ground truth input-into Gaussian features, and construct a hypergraph neural network to capture higher-order correlations across RGB and semantic information as well as between different frames. Hypergraphs use hyperedges to link multiple vertices, capturing complex relationships essential for cross-modal tasks. This higher-order structure addresses the limitations of NeRF and Gaussian methods, which lack the capacity for such advanced associations. This framework enables precise novel view synthesis and 2D semantic reconstruction without manual annotations, achieving state-of-the-art results for RGB and semantic tasks on room-scale scenes in the ScanNet and Replica datasets, while supporting real-time rendering speeds of 34 FPS.