Building the Next Generation Digital Maps Using a Fusion of 3D Computer Vision and Deep Learning

The past decade has seen a revolution in the ability of machines to automatically extract high level, semantic meaning from raw sensory data. This revolution has largely been brought about by the field of Deep Learning. In combination with an exponential growth in availability of compute and rapidly decreasing cost of sensing, this provides for a rather explosive combination when it comes to the field of digital mapping. The 3D Vision team of Apple Maps capitalizes on these trends to push the boundaries of mapping. By going deep into state of the art technologies like Machine Learning, 3D Computer Vision and large scale sensing & compute, we are able to redefine what the modern digital map is and how it is built. In the talk, we will walk you through our end to end philosophy for large scale mapping in the age of Deep Learning. With examples and visualizations, we take a look under the hood and explain how, over the last few years, our approach has led to a number of breakthrough products like Look Around, Visual Localization for geo referenced AR and most recently the new 3D Apple Maps.