Audio forgery has emerged as a significant security and forensic challenge, driven by rapid advances in generative artificial intelligence and the widespread availability of audio editing tools, which enable the creation of highly realistic manipulated speech with minimal technical expertise. Existing approaches predominantly operate at the file level, providing only coarse binary decisions without identifying when or where manipulation occurs. This study addresses fine-grained temporal localization through a unified frame-level localization framework. We introduce a controlled forgery generation framework derived from the TIMIT speech corpus, applying atomic, localized manipulations under strict temporal constraints and producing precise frame-level annotations across diverse manipulation types. Building on this dataset, we then propose a transform-agnostic localization-driven detection approach using temporal inconsistency modeling, enabling unified analysis across heterogeneous manipulations at frame-level resolution. To analyze forensic evidence, we present an evidence-stratified modeling paradigm comparing three complementary strategies: a handcrafted anomaly-based method, a deep localization model leveraging pretrained wav2vec 2.0 representations, and a hybrid approach combining both through confidence-aware fusion and temporal consistency reinforcement. A systematic experimental analysis evaluates the effects of representation adaptation, hybrid fusion, and manipulation type on detection and localization performance. Results show that handcrafted features are insufficient for reliable frame-level localization, while task-adapted wav2vec 2.0 achieves strong and consistent performance. The hybrid approach does not consistently improve frame-level accuracy but yields substantial gains in segment-level localization by enforcing temporal coherence. Per-transform analysis confirms robust performance across most manipulations, with deletion-based operations remaining the most challenging.

The rapid development of small unmanned aerial vehicles (UAVs) has introduced significant challenges in both civilian and defense security domains, such as public safety and protecting critical infrastructure. Detecting and classifying misused drones intruding into low-altitude, cluttered environments remains particularly challenging because of limitations in robust fusion across multiple sensing modalities. Although acoustic and visual fusion has been widely studied, its use for detecting intruding drones has not been sufficiently investigated. This paper proposes DroneAVFusion, a lightweight acoustic–visual fusion pipeline that integrates acoustic and visual sensing via a novel Late Attention Fusion (LAF) mechanism. The proposed system combines a Mel-spectrogram-based convolutional neural network for acoustic detection and optimized YOLO-based visual detection, followed by a computationally efficient fusion module that dynamically weights modalities using confidence-aware gating and temperature-scaled attention. A custom acoustic–visual dataset comprising three commercial drones and non-drone objects is collected and used to evaluate performance under realistic conditions. Experimental results show that DroneAVFusion achieves a $99.6$% F1 score on the validation set and perfect performance on the test set while running at an average of $30.40$ FPS with only $72,868$ parameters in the fusion module. Compared with single-modality baselines and existing fusion approaches, the proposed method demonstrates excellent robustness, making it well-suited for security applications.

Digital twin systems are becoming an important tool in intelligent transportation management, as they provide simulation-based environments for monitoring, analyzing, and predicting traffic behavior. However, the predictive performance of traffic digital twins is often limited by the quality and temporal consistency of sensor-level data generated from microscopic simulations. Most current calibration methods focus mainly on matching macroscopic traffic indicators, such as vehicle count and speed, without explicitly addressing the requirements of multi-horizon forecasting. This creates a gap between achieving realistic simulations and building reliable predictive models. This research proposes a forecasting-aware digital traffic twin framework that integrates microscopic SUMO simulation, controlled sensor-level observation modeling through geometric misalignment and noise injection, behavioral calibration, and deep temporal forecasting within a unified end-to-end structure. Unlike traditional calibration approaches, the proposed Genetic Algorithm (GA) reformulates calibration as a multi-step predictive optimization task. Simulation parameters are optimized by minimizing forecasting error produced by a lightweight proxy sequence model embedded within the calibration loop. In this way, calibration moves beyond simple statistical matching and instead emphasizes temporal learnability and forecasting stability, enabling the digital twin to generate traffic patterns more suitable for long-term prediction. Based on the calibrated traffic time series, both convolutional and recurrent deep learning models are evaluated under single-step and multi-step forecasting scenarios. To further examine generalizability, external validation is performed using the real-world PEMS-BAY dataset. The experimental findings demonstrate that forecasting-aware calibration reduces macroscopic traffic signal errors by around 50% for vehicle count and around 40% for average speed, improves temporal stability, and significantly enhances forecasting accuracy across both short-term and long-term horizons.