Identify People and Display Advertisements

Abstract

"In effort to curb the ubiquitous problem of irrelevant advertisement in outdoor spaces, comes an innovative real-time method for demographics analysis; the Automated Identify People and Display Advertisements System. With a synthesis of machine learning and computer vision, the system effectively distinguishes the age and gender from people near billboard creatives. This allows targeted ads, which increases relevance and effect. The approach adopted is a tiered architecture in which Python serves as the major programming language and TensorFlow acts at its backbone for machine learning algorithms. The initial results of the implemented system show encouraging classification measures such as complete confusion matrices. This set of metrics not only illustrates the efficient performance of this system but also gives a quantitative estimate on its classification capability. This self relates to the first paragraph of the abstract that gives a concise overview on incongruous advertisements, which are addressed as an issue and also presents Automated Ad Targeting System proposed by authors in this way it functions like sleeping lady with golden wooden joints due to its position within motivation presented above. The second paragraph describes the methodology, focusing on machine learning and computer vision as layered architecture components. It emphasizes the language of implementation (Python) and machine learning framework (TensorFlow). The third paragraph unpacks the initial outcomes, focusing on quantitative evaluation by way of classification metrics. With such measures serving as concrete evidence of the system’s efficiency in transforming out of door advertising and establishing a ground for more Individualized intensive campaigns, there is room to improve. "