To address the joint requirements of illegal parking detection and license plate recognition under complex traffic scenarios and limited edge-device resources, this study proposes RKF-YOLO, a lightweight dual-task model based on improved YOLOv11n that integrates Rep-CSP structural optimization, knowledge-transfer-enhanced training (KTET), and Focal-CIoU loss. Compared with YOLOv11n, RKF-YOLO reduces parameters and FLOPs by 38.2% and 38.1%, respectively, while improving $\mathrm{mAP}@0.5$ and $\mathrm{mAP}@0.5:0.95$ by 0.6 and 1.1 percentage points for parking detection; for plate detection, Focal-CIoU improves $\mathrm{mAP}@0.5:0.95$ by 1.3 percentage points and contributes to a recognition accuracy of 95.7%. The unified framework uses a shared backbone and task-oriented detection heads to support vehicle-level illegal parking detection and license-plate-oriented localization. Rep-CSP enhances multi-scale feature representation, asymmetric channel reduction with feature compensation reduces redundant computation, and KTET improves convergence through optimizer and learning-rate migration. Deployment on RK3588 achieves 59.5 FPS for parking detection and 95.1% recognition accuracy, demonstrating real-time performance and practical applicability on resource-constrained edge devices.

Accurate segmentation of retinal blood vessels is the core technical basis for computer-aided diagnosis of multiple fundus diseases, including diabetic ret

Medical Visual Question Answering (Med-VQA) integrates medical image understanding and natural-language reasoning, but strong benchmark results do not yet establish clinical reliability. This structured review focuses on representative Med-VQA models published from January 2022 through 30 September 2026, while retaining earlier methods and datasets only as historical and technical context. The revised review reports its information sources, search strategy, eligibility criteria, screening logic, and data-charting procedure. It compares classification-based and generation-based systems according to their answer-production mechanism, surveys ten public datasets, and critically interprets reported performance in light of heterogeneous data splits, preprocessing, answer vocabularies, and evaluation protocols. Particular attention is given to visual grounding, hallucination, domain generalization, experimental randomness, reproducibility, and the gap between benchmark accuracy and clinical validation. The review concludes that progress toward trustworthy Med-VQA requires evidence-grounded answers, uncertainty-aware evaluation, repeated-seed reporting, external validation, and sustained human oversight.