2 871 419 libros electrónicos en 110 idiomas
¿No le conviene? No hay problema. Puede devolverlo en un plazo de 30 días
No se equivocará con un vale de regalo. El destinatario puede elegir cualquier producto de nuestra oferta.
Política de devolución de 30 días
The goal of object recognition is to label objects fromimages and to estimate the poses of the labeled objects. The fieldof object recognition has seen tremendous progress with successfulapplications in some specific domains such as face recognition.However, the current state-of-the-art methods show unsatisfactoryresults for more general object domains in complex naturalenvironments with visual ambiguities. In this dissertation, we aimto enhance the object identification and categorization with theguide of visual context and graphical model. In this work, wepropose a general framework for the cooperative objectidentification and object categorization. Examplars used inidentification provide useful information of similarity incategorization. Conversely, novel objects are rejected inidentification but the proposed object categorization can label thenovel objects and segment them out for database update inidentification. This work can be helpful to the engineers inartificial intelligence and machine vision.
¡Hola! Soy Libroamiko, tu asesor de libros.
¿Cómo puedo ayudarte?