boundary-aware span extraction

This paper addresses the challenge of vast and structurally complex decision boundaries in DNA read mapping to reference genomes in next-generation sequencing (NGS). It investigates the boundary properties of naïve Bayes classifiers under graph-structured input spaces. To this end, the authors propose “neighborhood similarity” — a novel uncertainty measure that is both theoretically interpretable and universally computable, overcoming the reliance of conventional Bayesian confidence on model outputs. Leveraging graph-model-driven boundary analysis, neighborhood distribution statistics, and uncertainty quantification, the study reveals the high-dimensional complexity of decision boundaries and proves that the proposed measure simultaneously captures intrinsic Bayesian uncertainty. Moreover, it seamlessly extends to black-box classifiers lacking built-in confidence mechanisms. Empirically, neighborhood similarity significantly enhances classification interpretability and robustness, offering a principled framework for uncertainty-aware read mapping in NGS applications.

Existing video segmentation methods for long videos require predefined priors—such as thresholds, target segment counts, or length constraints—leading to inaccurate boundary localization. To address this, we propose a fully parameter-free video segmentation framework that automatically identifies semantically coherent segment boundaries without any human-specified hyperparameters. Our approach innovatively integrates the Minimum Description Length (MDL) principle with dynamic programming, enabling principled, data-driven boundary detection. It employs frame-level multimodal feature modeling and an optimized boundary search strategy to closely approximate the intrinsic structural semantics of real-world scenes. Evaluated on long-video summarization and retrieval-augmented video question answering, our method consistently outperforms state-of-the-art segmentation baselines. Downstream task performance improves significantly, demonstrating strong generalizability and practical utility across diverse video understanding applications.

研究通过引入多分割边界决策（MSBD）方法，减少模型调用次数，提高零样本页面流分割的效率和准确性。

This work addresses the unmodeled semantic constraints—namely, group continuity and partial ordering—in boundary labeling, aiming to automatically generate non-overlapping, semantically compliant label layouts under geometric and connectivity constraints. We provide the first formal proof that this problem is NP-hard under multiple hard constraints, including fixed endpoints, orthogonal edges, and minimum inter-label spacing. To tackle it, we propose a novel hybrid algorithmic framework with theoretical guarantees (a constant-factor approximation ratio) and practical efficiency, integrating integer linear programming, computational geometry optimization, greedy heuristics, and dynamic programming-based pruning. Evaluated on standard benchmarks, our method reduces crossing edges by 37% compared to state-of-the-art approaches, achieves a label placement success rate of 98.2%, and maintains an average runtime below 10 milliseconds.

Subspace clustering in high-dimensional data often yields multiple semantically distinct subspaces, yet existing methods require manual specification of both the number of subspaces and the number of clusters within each—rendering them parameter-sensitive and poorly interpretable. This paper proposes an automatic, non-redundant multi-subspace clustering framework. First, it introduces the Minimum Description Length (MDL) principle to non-redundant clustering, enabling joint, adaptive inference of both the optimal number of subspaces and the cluster count per subspace. Second, it designs a split-merge-based greedy search strategy coupled with a subspace-level outlier encoding mechanism, allowing simultaneous outlier detection. Evaluated on multiple benchmark datasets, the method achieves competitive accuracy against state-of-the-art approaches while significantly improving parameter robustness, model interpretability, and practical applicability.

This work addresses document-level conspiracy theory detection by proposing a joint framework that integrates multi-label span classification with sequence classification. For extracting conspiracy-related markers—such as roles and actions—the approach formulates the task as boundary-aware multi-label span classification, incorporating IoU-based positive labeling, hard negative sampling, and an inclusion-aware non-maximum suppression strategy, while distinguishing between entity-like and abstract roles. Document-level classification is performed using a RoBERTa model enhanced with label smoothing. Evaluated on SemEval-2026 Task 10, the method achieves 7th place in Subtask 1 (macro F1 = 0.2251) and 11th place in Subtask 2 (weighted F1 = 0.7694), demonstrating the effectiveness of the proposed techniques.

This work addresses the challenge of automatically determining the optimal number of clusters for high-dimensional, complex data—such as large-scale images—without any prior information. It proposes a novel method that circumvents assumptions about data distribution and does not require complete clustering results. The approach reformulates cluster number estimation as a dynamic comparison of positional relationships among cluster centers, introducing for the first time a sample confidence filtering mechanism to exclude low-confidence boundary samples. By integrating bipartite graph modeling with a pairwise center-matching strategy, the method achieves robust performance. Extensive experiments on challenging benchmarks, including CIFAR-10 and STL-10, demonstrate its significant superiority over current state-of-the-art techniques, highlighting enhanced robustness and adaptability.

This study addresses the extreme class imbalance in whiteboard stroke binary segmentation, where foreground pixels constitute only 1.79% on average, rendering conventional region-based metrics inadequate for evaluating fine-stroke performance. To this end, the authors propose a comprehensive evaluation protocol that integrates region- and boundary-based metrics, fairness analysis across core and fine-stroke subsets, and robustness statistics from multi-run training. Using a DeepLabV3-MobileNetV3 architecture, five loss functions are systematically compared. Results show that overlap-aware losses (e.g., Dice+Focal) improve F1 by over 20 points (0.663 vs. 0.438) compared to cross-entropy, while Tversky loss significantly outperforms the Sauvola method in worst-case F1 (0.565 vs. 0.452). Increasing training resolution further boosts F1 by up to 12.7 points. This work is the first to incorporate non-parametric significance testing and worst-case evaluation, uncovering hidden trade-offs among loss functions in fine-grained segmentation tasks.

This study addresses the degradation in recognition performance caused by occlusion-induced loss of boundary information in images. Inspired by the boundary completion mechanism in the visual cortex, this work proposes BorderNet, a novel convolutional neural network architecture that, for the first time, integrates a mathematical model of this biological mechanism into deep learning. By incorporating biologically inspired convolutional filters and an occlusion-robust training strategy, BorderNet effectively enhances the model’s ability to perceive and reconstruct missing boundaries. Experimental results on three occluded benchmark datasets—MNIST, Fashion-MNIST, and EMNIST—demonstrate that BorderNet significantly outperforms existing baseline methods, with particularly notable improvements in classification accuracy under severe occlusion conditions.

本文提出POSPAN框架，通过结合跨度长度分布与位置约束分布来改进语言模型预训练中的跨度掩码策略，增强了模型性能。