cs papers | Gist.Science

Distilling Facial Emotional Cues into EEG Representations for Robust Emotion Recognition in Mental Health Monitoring

The paper proposes Cross-Modality Enhanced Distillation (C-MED), a framework that transfers multimodal EEG-facial knowledge to a deployable EEG-only student model via contrastive alignment and bilinear interaction, achieving robust, subject-independent emotion recognition on the SEED-VII and DEAP benchmarks without requiring synchronized facial data at inference.

Majid Sepahvand, Hiba Muhammed Hussein, Maytham N. Meqdad2026-09-10

💻 computer science

Dynamic ensemble approach for multi-class classification based on neighborhood rough sets and sequential three-way decisions

This paper proposes EM-S3WD, a dynamic ensemble framework that integrates neighborhood rough sets with sequential three-way decisions and a conditional dynamic integration strategy to overcome the limitations of fixed reference tuples and binary constraints in Xu et al.'s original model, thereby achieving adaptive and competitive multi-class classification performance.

Wenyan Xu, Qiang Chen, Yangyang Guo2026-09-10

💻 computer science

Artificial Intelligence Driven Grading and Personalised Feedback in Higher Education Assessment

This systematic literature review synthesizes evidence from 2020 to 2026 to demonstrate that while AI-driven grading and personalized feedback significantly enhance efficiency, consistency, and scalability in higher education assessment, their responsible implementation requires robust ethical safeguards and human oversight to address challenges such as algorithmic bias, data privacy, and academic integrity.

Mziwendoda Cyprian Madwe2026-09-10

💻 computer science

TAG-lite: a pre-training task-affinity workflow for multi-task ADMET learning with disjoint datasets

The paper introduces TAG-lite, a gradient-based pre-training workflow that estimates task affinity between disjoint ADMET datasets to effectively group endpoints and improve multi-task learning performance even when compound overlap is minimal.

Seoyoon Koo, Sungmin Ko, Seung-Yong Seong2026-09-10

💻 computer science

Genre Classification of Saint Yared Qum Zema Using a Convolutional Neural Network

This study presents a custom Convolutional Neural Network model that achieves 88% testing accuracy in classifying the three genres of Saint Yared's Ethiopian sacred music (Qum Zema) by converting 1,555 audio segments into spectrogram images, outperforming established architectures like ResNet, VGGNet, and AlexNet.

Birku Litgeb Aschenek, Alehegn Ayanaw Getaneh, Getaneh Awoke2026-09-10

💻 computer science

Hierarchical Compositional Hypergraphs Encode Document Structure for Classification

This paper introduces a Hierarchical Compositional Hypergraph (HCH) that encodes document structure through ordered token, sentence, and paragraph layers, demonstrating that combining these structural features with standard TF–IDF yields statistically significant improvements in text classification accuracy and macro-F1 over lexical baselines alone.

Madjid Eshaghi Gordji, Mohamadali Berahman2026-09-10

💻 computer science

CauCity-WM: An Intervention-Consistent Generative World Model for Counterfactual Urban Dynamics and Risk-Aware Decision Making

The paper introduces CauCity-WM, a causal generative world model that unifies urban forecasting, counterfactual estimation, and risk-aware decision making by employing a twin-world structural diffusion process to generate consistent factual and counterfactual futures through abducted exogenous noise.

Kai Ma, Miao Yu2026-09-10

💻 computer science

Interpretable ensemble machine learning identifies player-specific performance indicators in elite tennis

This study demonstrates that an interpretable ensemble machine learning workflow applied to Novak Djokovic's Grand Slam data not only achieves high predictive accuracy for set outcomes but also reveals player-specific performance indicators, such as backhand returns, that are often overlooked by individual models.

Yifeng Li, Weixiang Jiao, Meiyi ZouZhu, Yawen Zheng, Wenjie Lu, Yihang Kong, Xingyun Li2026-09-10

💻 computer science

VectorLLM++: A Unified Next-Token-Prediction MLLM for Dense Remote Sensing Perception and Vector Mapping

VectorLLM++ is a unified next-token-prediction multimodal large language model that overcomes data scarcity and the lack of structured vector outputs by introducing novel tokenizers for coordinates and masks, enabling it to outperform specialized models across seven diverse remote sensing perception and vector mapping tasks.

Shunping Ji2026-09-10

💻 computer science

Native Byzantine-Robust Aggregation for Trustworthy Federated Learning: A C++20 Evaluation of Krum, Multi-Krum, Trimmed Mean, and Coordinate-wise Median

This paper presents and evaluates a high-performance C++20 implementation of Byzantine-robust aggregation algorithms (Krum, Multi-Krum, Trimmed Mean, and Median) for Federated Learning, demonstrating through rigorous correctness checks and benchmarks that combining explicit Byzantine assumptions with numerically defensive native systems design significantly accelerates aggregation while maintaining robustness against adversarial updates.

Md Shahanur Islam Shagor2026-09-10

← Previous Next →