cs papers | Gist.Science

External Validation and Interpretability Analysis of a Deep Learning Model for Forensic Age Estimation Across Brazilian and Romanian Populations

This study demonstrates that while deep learning models can capture biological aging patterns for forensic age estimation, their reliability is significantly compromised by population and hardware shifts, underscoring the critical need for external validation and interpretability analysis to detect non-anatomical shortcuts before forensic application.

Laura Meloni, Liviu-Mihai Iacob, Sorana Eftimie, Raluca Roman2026-08-04

💻 computer science

SPARE - STARE: Securing Progressive Web Apps - Trusted Web Activity Against Unauthorized Replications

This paper proposes and evaluates "SPARE-STARE," a query-parameter-based security framework that uses encryption and unique identifiers to bind Progressive Web Apps to Trusted Web Activities, effectively mitigating unauthorized replication threats while overcoming communication limitations and reducing development complexity.

Sajib Talukder, Nur Imtiazul Haque, Hisham Kholidy2026-08-04

💻 computer science

Loss-Aware Attention-Based Dual Feature Selection for Handling Imbalanced IoT Intrusion Detection

This paper proposes the DFLA-IDS framework, which combines global feature selection using MIC and mRMR with a loss-aware attention transformer to effectively address feature redundancy and class imbalance, achieving superior intrusion detection performance on the UNSW-NB15 dataset compared to Deep BiLSTM.

Majid Altuwairiqi Altuwairiqi2026-08-04

💻 computer science

IFF-HVT: Informative Feature Fusion in Hybrid Vision Transformers for Image Classification

This paper introduces IFF-HVT, a hybrid architecture that integrates a ResNet-50 backbone with transformer-based context modeling and an informative feature fusion module to effectively balance local and global feature representation, achieving significant accuracy improvements over baseline models on mid-scale image classification datasets.

Ahsan Umar2026-08-04

💻 computer science

Comparative Evaluation of Machine Learning and Deep Learning Models for Targeted Metaphor Detection

This paper evaluates various machine learning and deep learning models for targeted metaphor detection on a dataset of seven target words, finding that a hybrid LDA-Sentence-BERT-Random Forest approach outperforms other methods, including a consensus ensemble, by achieving 84% accuracy through the effective combination of contextual sentence embeddings and topic-level information.

Muhammad Hassam Aslam Khan, Ninad Deshpande2026-08-04

💻 computer science

Spatial Fairness in Taxi Demand Prediction: A Hybrid Approach Leveraging District Profiles

This paper proposes a hybrid taxi demand prediction framework that integrates Non-negative Matrix Factorization with White Shark Optimization for data imputation, LSTM for temporal modeling, and a fairness-aware regularization mechanism based on Relative Hardship Proximity to minimize prediction disparities across urban districts while maintaining high accuracy.

Ghazalak Eslami, Foad Ghaderi2026-08-04

💻 computer science

Eco Track: AI- and IoT-Powered E-Waste Segregation and Awareness for Sustainable Recycling

This paper presents "Eco Track," a low-cost, AI- and IoT-powered prototype utilizing a YOLOv8-M model and motorized actuators to achieve automated, safety-aware segregation of small e-waste components, demonstrating high precision and mAP in laboratory settings while addressing practical edge deployment constraints.

R Rashmi, Konakalla Meenakshi, Medhini Katti, M.R. Madhura2026-08-04

💻 computer science

EELTRA-Net: An Intelligent Hybrid Framework for Energy-Efficient and Lifetime-Aware Routing in Vehicular Ad Hoc Networks

EELTRA-Net is an intelligent hybrid framework that integrates Temporal Convolutional Units, Gated Recurrent Units, and an Attention-Driven Feedforward Network to optimize energy efficiency, reliability, and network lifetime in Vehicular Ad Hoc Networks, achieving superior performance metrics such as a 98.9% Packet Delivery Ratio compared to existing baseline models.

Krishna Komaram, Nagarjuna Karyemsetty2026-08-03

💻 computer science

AMS-Net: Lightweight Self-Attention Enhanced MobileNetV3 for Real-Time Multimedia Classification on Edge Devices

This paper proposes AMS-Net, a lightweight MobileNetV3-Small architecture enhanced with a novel Low-complexity Spatial Self-Attention Module and a dynamic transfer learning strategy, which achieves superior real-time classification accuracy and efficiency on resource-constrained edge devices.

Shitong Wang, Jiaquan Shen, Jiaxi Yang2026-08-03

💻 computer science

How Do Algorithms Interact in Scientific Research? Large-Scale Semantic Relation Identification and Analysis Based on Full-Text Papers

This study constructs a framework to identify and analyze semantic relations among algorithm entities in NLP full-text papers, revealing that SciBERT excels in relation extraction while the field exhibits expanding, deepening, and rapidly evolving interaction networks dominated by comparative relationships.

Yuzhuo Wang, Yi Zhao, Wenqing Wu, Heng Zhang2026-08-03

← Previous Next →