An increasing amount of information is being conveyed via images, with mobile photographs being a particularly notable case. While image metadata, such as the text linking to an image in Google image search, is a common input to machine learning tasks, the content of the image itself is used far less frequently. Whether picking your best vacation photos, recommending similar-looking images to use in a talk or searching for hidden cats, learning tasks can do better if the actual pixels are used as input.
However, The choice of features determines the quality of result as much as the choice of machine learning algorithm and using the pixels directly often yields the poor results. Higher-level image features, such as face detection, histograms and color statistics like hue binning, provide significantly better performance. While advantageous, these features force the developer to choose from a vast number to accurately capture the details of their problem domain, a challenging task. This talk covers classes of simple image features and how to employ them in machine learning algorithms and focuses on providing basic domain knowledge in imaging/computer vision to developers already familiar with machine learning.
The outline is as follows: We begin with an overview of common image features and discuss potential applications for each. Common features include examples from computer vision such as blob identification, face detection and edge statistics as well as from image statistics such as intensity histograms, Fourier properties and color statistics such as hue binning. Next, we present how to generate the features with python imaging libraries. Finally, we discuss approaches to converting complex image features into a series of scalars for the input vector of an ML algorithm that best represent the problem domain.
Matthew Trentacoste is a Computer Scientist at Adobe Systems, where he develops mobile photography platforms. He's interested in computational photography, networked cameras and how large collections of images can enable new forms of creative imaging.
His background is a mix of image processing, human visual perception, data analysis and mobile platforms. Previously, he's worked or consulted at companies large and small, including Dolby, Pocket Pixels, BrightSide Technologies and Mobify. He holds a degree in Computer Science from Carnegie Mellon and with a Ph.D. and M.Sc. from the University of British Columbia, both in Computer Science.
What is PyData?
PyData.org is the home for all things related to the use of Python in data management and analysis. This site aims to make open source data science tools easily accessible by listing the links in one location. If you would like to submit a download link or any items to be listed in PyData News, please let us know at: firstname.lastname@example.org
PyData conferences are a gathering of users and developers of data analysis tools in Python. The goals are to provide Python enthusiasts a place to share ideas and learn from each other about how best to apply the language and tools to ever-evolving challenges in the vast realm of data management, processing, analytics, and visualization.
We aim to be an accessible, community-driven conference, with tutorials for novices, advanced topical workshops for practitioners, and opportunities for package developers and users to meet in person.
A major goal of PyData events and conferences is to provide a venue for users across all the various domains of data analysis to share their experiences and their techniques, as well as highlight the triumphs and potential pitfalls of using Python for certain kinds of problems.
PyData is organized by NumFOCUS with the generous help and support of our sponsors. Proceeds from PyData are donated to NumFOCUS and used for the continued development of the open-source tools used by data scientists If you would like to volunteer to be a part of the PyData team contact us at: email@example.com