We study the task of panoptic symbol spotting, which involves identifying both individual instances of countable \textit{things} and the semantic regions of uncountable \textit{stuff} in computer-aided design (CAD) drawings composed of vector graphical primitives. Existing methods typically rely on image rasterization, graph construction, or point-based representation, but these approaches often suffer from high computational costs, limited generality, and loss of geometric structural information. In this paper, we propose \textit{VecFormer}, a novel method that addresses these challenges through \textit{line-based representation} of primitives. This design preserves the geometric continuity of the original primitive, enabling more accurate shape representation while maintaining a computation-friendly structure, making it well-suited for vector graphic understanding tasks. To further enhance prediction reliability, we introduce a \textit{Branch Fusion Refinement} module that effectively integrates instance and semantic predictions, resolving their inconsistencies for more coherent panoptic outputs. Extensive experiments demonstrate that our method establishes a new state-of-the-art, achieving 91.1 PQ, with Stuff-PQ improved by 9.6 and 21.2 points over the second-best results under settings with and without prior information, respectively—highlighting the strong potential of line-based representation as a foundation for vector graphic understanding.

Spatial-temporal action detection is a vital part of video understanding. Current spatial-temporal action detection methods mostly use an object detector to obtain person candidates and classify these person candidates into different action categories. So-called two-stage methods are heavy and hard to apply in real-world applications. Some existing methods build one-stage pipelines, But a large performance drop exists with the vanilla one-stage pipeline and extra classification modules are needed to achieve comparable performance. In this paper, we explore a simple and effective pipeline to build a strong one-stage spatial-temporal action detector. The pipeline is composed by two parts: one is a simple end-to-end spatial-temporal action detector. The proposed end-to-end detector has minor architecture changes to current proposal-based detectors and does not add extra action classification modules. The other part is a novel labeling strategy to utilize unlabeled frames in sparse annotated data. We named our model as SE-STAD. The proposed SE-STAD achieves around 2% mAP boost and around 80% FLOPs reduction. Our code will be released at https://github.com/4paradigm-CV/SE-STAD

This paper investigates low-complexity multi-user detectors for massive MIMO systems in practical channels. Beam domain message passing algorithm on the sparse-graph is proposed by exploiting the channel sparsity in beam domain, which has much lower computational complexity compared to the dense-graph based detections. As the massive MIMO channels are usually not i.i.d Gaussian distribution and central limit theorem (CLT) will no longer apply, conventional approximate message passing detector (AMP) may diverge. To improve the performance of AMP in practical channels, we further propose energy spread transform (EST) based massive MIMO transmission, which guarantee that the equivalent channels are i.i.d. distributed. Furthermore, by deriving the state evolution (SE) equations, the convergence analysis of the proposed detector is also provided. Simulation results show that, the proposed detector outperforms the conventional AMP and minimum mean square error (MMSE) detectors, and has fast convergence rate.

This paper investigates the pilot design and channel estimation issue for massive MIMO-OFDM uplink transmission. By exploiting the channel sparsity in the beam-delay domain, the channel estimation issue for multiuser uplink transmission is formed into a sparse compressed sensing (CS) model. We propose a cyclic shift pilot scheme, which satisfies Restricted Isometry Property under the CS frameworks. As the sparsity is efficiently utilized, the proposed scheme significantly reduces the pilot overhead, especially when the number of users becomes large. Furthermore, based on approximate message passing (AMP) algorithm, an approximate Bayesian inference can be conducted for channel estimation with low complexity. Simulation results show that the proposed AMP-based algorithm with the pilot scheme outperforms several existing baselines, especially when the BS is equipped with large antenna arrays.