2608 papers
This paper presents DORI, a distributed interface that ensures resilient multi-source event synchronization in cloud-native systems by applying reconciliation-by-design principles to detect inconsistencies, validate blockchain events, and provide idempotent corrections before atomically delivering them to downstream microservices.
This paper proposes a speed and morphology-conditioned Fourier-based gait synthesis method that converts variable-duration joint angle data into fixed-dimensional frequency coefficients for robust reconstruction, demonstrating superior performance over existing baselines in both signal fidelity and downstream reinforcement learning tasks for bipedal locomotion.
This study proposes and validates a refined hybrid retrieval-augmented generation framework, utilizing a sparse-dense-reranking multi-stage pipeline with optimized chunking and cascading thresholds, to significantly improve retrieval accuracy and semantic alignment for high-logic-density campus policy texts.
This review synthesizes evidence from representational, behavioral, and mechanistic perspectives on subject-verb agreement, negation, and compositional generalization in pretrained Transformers to reveal distinct patterns of convergence and fragmentation, ultimately proposing an accessibility-and-use diagnostic to clarify current limitations and guide future cross-level research.
This paper presents an Edge AI and IoT framework that integrates YOLOv8s-based cattle detection with MQTT-driven communication and automated acoustic deterrence to enable real-time, autonomous livestock intrusion monitoring and farm protection.
This study demonstrates that synthetic rewriting acts as a quality multiplier rather than a substitute for data curation in Portuguese continued pretraining, significantly boosting performance only when applied to high-quality source data and primarily at larger model scales.
This paper introduces a benchmark framework for evaluating five families of unsupervised trajectory vectorization techniques across synthetic and real-world datasets, demonstrating their effectiveness in clustering and identifying specific trajectory characteristics such as motion dynamics and geometric patterns.
This survey provides a comprehensive overview of tweet summarization methods, evaluating extractive and abstractive approaches, pre-processing strategies for noisy short texts, and emerging trends like large language models, while identifying key challenges and future research directions.
This study demonstrates that under controlled PlantVillage conditions, classical machine learning models utilizing handcrafted color-texture features achieve performance statistically comparable to MobileNetV2 transfer learning for tomato leaf disease classification, challenging the assumption that deep learning is always superior for this task.
This paper proposes a reinforcement learning supervisor framework that dynamically weights multiple performance metrics to select the optimal trading policy from a heterogeneous set of agents, demonstrating superior risk-adjusted returns in nonstationary cryptocurrency markets compared to individual agents and fixed strategies.