Founder, CEO, Umajin. Palmerston North, New Zealand.

“Sparsification“

18 February 2025

Abstract

This approach to 3D time series data reduction allows for a smaller node counts by 10,000 to 1,000,000 times over voxels.  This also allows for multiple pre computed connectivity graphs to be stored to make simulation and analysis of time series data sets much faster. To further reduce workloads spatial partitioning can be used to support variable density.  By assessing the change on a region by region basis it is also re-computed only when certain spatial thresholds are reached.

Bio

David’s team develops the Umajin platform which is designed as a cloud and server technology with support for user interaction, 3D machine vision, AR, IoT and AI. The Umajin runtime has shipped over 30 million instances shipped with partners like NTT, PwC, Dell, HP, Intel and Sony.

---

MW25 Slides

MW25 Videos

MW25 Q & A

Video 2024

David is a returning speaker. Please refer to his 2023 talk “High performance object detection and classification on smartphone class hardware” and watch his interview that year.