AI-based targeted advertising system

Abstract

<span>The most common technology used in targeted advertising is facial recognition and vehicle recognition. Even though there are existing systems serving for the targeting purposes, most propose limited functionalities and the system performance is normally unknown. This paper presents an intelligent targeted advertising system with multiple functionalities, namely facial recognition for gender and age, vehicle recognition, and multiple object detection. The main purpose is to improve the effectiveness of outdoor advertising through biometrics approaches and machine learning technology. Machine learning algorithms are implemented for higher recognition accuracy and hence achieved better targeted advertising effect.</span>

Gender recognition from unconstrained selfie images: a convolutional neural network approach

“…This is in contrast with other conventional algorithms that rely on selective pre-processed hand-crafted features. Furthermore, the availability of enormous, labelled archives of data, i.e., ImageNet, enabled CNNs to learn and infer valuable features' formations, which rendered a non-precedent breakthrough in various pattern recognition tasks that even exceeded human performance [2]. In addition, using a bit of network reshaping, these inferred representations and/or features could be surprisingly reused and applied to a newly different problem.…”

“…Automated soft biometrics identification has attracted a great deal of attention in the past era. This was primarily related to the uprising dependence on surveillance systems that produces enormous volumes of data that need to be examined [1], [2] in an off-line manner. Furthermore, it is possible to obtain soft biometrics without subject participation from low quality videos/images, making them highly valuable [2].…”

“…This was primarily related to the uprising dependence on surveillance systems that produces enormous volumes of data that need to be examined [1], [2] in an off-line manner. Furthermore, it is possible to obtain soft biometrics without subject participation from low quality videos/images, making them highly valuable [2]. In addition, people normally utilize these soft biometric to identify each other [1].…”

Kidney stones detection based on deep learning and discrete wavelet transform

“…discrete first chebyshev wavelet transformation (DFCWT) constructed from (1) reached to DFCWT of the following equations the wavelets are divided into scaling function 𝜕(𝑡)in 𝑉 𝑛 space and first chebyshev wavelet 𝛿(𝑥) in wavelet space 𝑊 𝑛 𝛿 𝑢,𝑣 (𝑡) = 𝛿 𝑡,𝑢,𝑣,𝑘 𝑘 = 1,2 𝑢 = 1,2, … , 2 𝑘−1 , v is the order of first chebyshev polynomials, the transform DFCWT 𝑒 = 2 −(𝑘−1) and 𝑓 = 2 −(𝑘+1) (2𝑢 − 1)for transform, 𝑥 = 2 −(𝑘−1) (2 𝑘 𝑡) in (1).…”

“…The role of artificial intelligence in the medical field as well as in many other fields is shown through deep learning in general and the convolutional neural network in particular, there has been interest in artificial intelligence and deep learning recently due to the large and massive amounts of data that need to be examined and knowledge of the sample being examined [1], [2]. Diab et al [3] deep learning and convolutional neural network training were used to detect brain tumors, especially ResNet50 was used, with an accuracy of 99.8% with an error of 0.005.…”