Computer vision and image processing are some of the hardest problems in machine learning. Challenges involving categorization of images and object recognition usually involve a lot of pre processing and transformations in order to extract meaningful signal for which to accomplish these tasks. Enter deep learning. With recent advancements in neural networks and improved hardware, neural networks have been used for state of the art performance on many kinds of image processing tasks.
Come learn about some of the tricks used to feed images in to a neural network, and some of the neat transformations neural networks automatically perform to allow you do focus on your target task.
At this Meetup, Zipfian Academy machine learning instructor and creator of Deeplearning4j Adam Gibson will give an introduction to his distributed, open-source deep learning framework. This will be the first of a monthly series by Adam on Deeplearning4j, the first commercial-grade deep learning library written in Java.
This Meetup is an overview of the techniques used for image processing and how to begin interpreting neural network output and know whether you are on the right track.
The talk series is intended for those with a working knowledge of machine learning (i.e., scikit-learn).
Sign up if you're ready to commit to a series of workshops that will show you how to apply DL4J to machine vision, text and times series analysis, and sound recognition.
About Adam Gibson:
Adam the co-founder of Skymind, a business intelligence and enterprise software firm, and blix.io, a machine learning and NLP startup that helps public relations professionals to monitor coverage and avoid disasters. He is a machine-learning instructor at Zipfian Academy. Before, Adam did Data Mining at Mashape, ran his own web development consultancy, and was a Dev Ops engineer.