Buscador :
Volver al Menú
| : /
Vote:
Resultados:
0 Votos
NOVIEMBRE 2026 - Volumen: 101 - Páginas: 541-547
¿Le interesa este artículo? Puede comprar el artículo a través de la plataforma de pago de PayPal o tarjeta de crédito (VISA, MasterCard,...) por 20 €.
Abstract: With the rapid development of the tourism industry, the demand for real-time monitoring and accurate personnel identification in scenic spot safety management is becoming increasingly urgent. Facial recognition technology has become the core means of scenic spot monitoring due to its high efficiency. However, traditional facial recognition methods face significant bottlenecks in complex monitoring scenarios in scenic areas. Therefore, the study proposes a facial recognition monitoring method based on an improved Residual Network 50 and Transformer. This method can improve the performance of facial recognition in surveillance systems in scenic areas and ensure tourism safety. This method enhances the model's ability to handle complex pose variations and occlusion issues by introducing coordinate attention, conditionally parameterized convolution, a spatial gating mechanism, and a squeeze-and-excitation module. The experimental results indicated that, in the CFP dataset, the feature extraction accuracy of this method reached up to 99.3%, with an equal error rate as low as 0.9%. In summary, an effective solution for facial recognition technology in scenic area monitoring is provided by the improved residual network 50-Transformer, and significant application value is held by it.Keywords: Tourism safety; Facial recognition; ResNet50; Transformer; Coordinate attention; SAE
Compártenos:
© Revista de Ingeniería Dyna 2025 - UK Zhende Publishing Limited
Dirección: Unit 7 Wilsons Business Park, Manchester M40 8WN United Kingdom
Email: office@revistadyna.com
Regístrese en un paso con su email y podrá personalizar sus preferencias mediante su perfil
Nombre: *
Apellido 1: *
Apellido 2:
Email: *