Comparative analysis of modern online tools for object recognition in images
Abstract
There are many practical tasks that require object recognition in images. Various approaches are used to solve this task, typically involving neural networks [1]. These networks have been pre-trained on a specific set of reference images, with corresponding objects assigned to specific classes.
In this work, an analysis of 4 online tools providing functionality for object recognition in images has been conducted. The tools are as follows: Amazon Rekognition, Google Cloud Vision, Microsoft Azure AI Vision Studio, and Imagga. To investigate and assess their performance, the graphical interface of each tool’s website was used. It is worth adding that these services allow users to interact with them both through their graphical interfaces and via API.
References
Krizhevsky, A., Sutskever, I., Hinton, G. E. (2017). Imagenet classification with deep convolutional neural networks. Communications of the ACM, 60(6), 84–90.
Rachev, S. T., Stoyanov, S., Fabozzi, F. J. (2008). Advanced stochastic models, risk assessment, and portfolio optimization: The ideal risk, uncertainty, and performance measures. Optimization (p. 382).
