cs papers | Gist.Science

Pretraining Transfer, Adaptation Dependence, and Dense Evaluation in Low-Label Cell Instance Segmentation: A Controlled Study

This controlled study demonstrates that rankings of pretraining strategies for low-label cell instance segmentation are not inherent to the backbone weights alone but critically depend on the specific transfer-and-adaptation protocols used, while also highlighting that standard evaluation caps significantly underestimate performance on dense microscopy images.

Yinghuan Li2026-07-28

💻 computer science

Meta-Analysis of Feature Representation Techniques for Fake News Detection: Evidence from Publicly Available Benchmark Datasets

This meta-analysis of publicly available benchmark datasets demonstrates that hybrid feature representations combined with ensemble learning approaches offer superior predictive stability and effectiveness for detecting fake news compared to individual feature extraction methods and classifiers.

Kazi Abdul Mannan, Tasima Tahsin Diya2026-07-28

💻 computer science

A Modular Image Manipulation Localization Framework using a Dual-Stream Classifier and Conditional Diffusion Models

This paper proposes a novel two-stage modular framework for image manipulation localization that first classifies forgery types using a dual-stream classifier fusing RGB and SRM features, then employs specialized conditional diffusion models with a hybrid loss function to generate precise manipulation masks, achieving competitive performance on benchmark datasets.

Zohaib Hamdule, Venkatanareshbabu Kuppili2026-07-28

💻 computer science

Active Image Manipulation Detection and Localization using TransUNet for Fragile Watermarking

This paper proposes a novel Active Image Manipulation Detection and Localization (AIMDL) framework that leverages TransUNet for robust watermark generation and recovery, combined with a multi-head UNet forensic module, to achieve high-accuracy tampering detection and pixel-level localization while dynamically balancing watermark transparency and robustness.

Zohaib Hamdule, Venkatanareshbabu Kuppili2026-07-28

💻 computer science

StrokePlan-RNN: Predicting Ordered Drawing Procedures from Face Line Art

This paper introduces StrokePlan-RNN, an image-conditioned autoregressive model that successfully predicts ordered stroke sequences from static face line art by training on a newly constructed dataset, achieving high geometric accuracy while highlighting the need for improved procedural evaluation metrics.

Gameil S. H. Ali, Entesar Altaheri2026-07-28

💻 computer science

Adaptive Training Controller (ATC): An Autotuning Framework for CNN-Based Plastic Waste Classification Using the PLASTIC Dataset

This paper addresses the challenge of plastic waste classification by introducing a new domain-specific PLASTIC dataset and proposing the Adaptive Training Controller (ATC), an architecture-independent autotuning framework that outperforms existing tools like Optuna by achieving up to 94.89% accuracy with ResNet101 while reducing computational resource requirements.

SHYAMASREE KARMAKAR, Sukanya Saha, Sunita Roy, Ranjan Mehera, Rajat Kumar Pal2026-07-28

💻 computer science

Can a Large Language Model Serve as the Missing Second Reviewer? Prominence Bias, Retrieval Effects, and a Confidence-Fabrication Gap in an Empirical Evaluation of Two Published Meta-Analyses

This empirical evaluation demonstrates that while Large Language Models can identify real errors missed by human reviewers and significantly improve study recall when equipped with retrieval tools, they remain unreliable substitutes for human verification due to persistent issues like prominence bias, cross-paper conflation, and the fabrication of non-existent inconsistencies.

Paul Fontelo2026-07-28

💻 computer science

Cyber-Physical Orchestration of Emergency Department Services Through a Bidirectional Digital Twin and a Deep Reinforcement Learning Agent: Architectural Design, In SilicoValidation, and Critical Analysis of Performance Under Demand Stress

This paper presents and validates a cyber-physical architecture integrating a bidirectional Digital Twin, a Deep Reinforcement Learning agent, and an Explainable AI layer to optimize emergency department dispatching, demonstrating significant reductions in peak occupancy under high-demand stress while critically exposing and addressing specific instrumentation artifacts in the validation process.

Joseph Javier Sánchez Acuña, Mario Anzures García2026-07-28

💻 computer science

The Machine Proposes. The Proof Disposes: Neuro-Symbolic Synthesis of Formally Verified Markov Usage Models from Natural Language Requirements

This paper introduces Neuro-Symbolic MBST, a framework that automates the synthesis of formally verified Markov usage models from natural language requirements by integrating L* learning, grammar-constrained LLMs, and convex optimization, thereby achieving high-fidelity fault detection and coverage that significantly outperforms pure-neural baselines while eliminating manual modeling bottlenecks for safety-critical systems.

Nathan Ginting2026-07-28

💻 computer science

Interactive Query based Abnormal Events Synopsis Generation in Surveillance Video

This paper proposes an interactive query-based algorithm for generating abnormal event synopses in surveillance videos that utilizes a rule-based classifier to handle complex user queries and introduces an "improved overlapping ratio" metric for evaluation, demonstrating superior accuracy and quality over existing methods on the PETS09 dataset.

Judi Vennila Thangaswamy, Balamurugan Vaniappan2026-07-28

← Previous Next →