Building upon the foundational understanding of how deep learning is applied to image classification, Alison Lowndes leads a hands-on exploration of approaches to the challenging problem of detecting if an object of interest is present within an image and, if so, recognizing its precise location within the image. You’ll learn how to apply cutting-edge object detection networks trained using NVIDIA DIGITS on a challenging real-world dataset.
Along the way, Alison walks you through testing three different approaches to deploying a trained DNN for inference: directly using inference functionality within a deep learning framework, in this case DIGITS and Caffe; integrating inference within a custom application by using a deep learning framework API, again using Caffe but this time through its Python API; and using the NVIDIA TensorRT, which will automatically create an optimized inference runtime from a trained Caffe model and network description file. You’ll discover the role batch size plays in inference performance and various optimizations that can be made in the inference process and explore inference for a variety of different DNN architectures.
Alison Lowndes is a solution architect and community manager at NVIDIA. Alison has 25+ years in international project management and entrepreneurship with two decades spent within the internet arena. In her spare time, she is a founder trustee of a global volunteering network. A very recent graduate in artificial intelligence at the University of Leeds, where she where she completed a thorough empirical study of deep learning, specifically with GPU technology, covering the entire history and technical aspects of GPGPU with underlying mathematics, Alison combines technical and theoretical computer science with a physics background.
©2017, O’Reilly UK Ltd • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • email@example.com