Machine Learning (ML)

Detecting “Action” and “Cut” in Archival Footage Using a Multi-model Computer Vision and Audio Approach with Azure Cognitive Services

Nile Wilson September 27, 2021 Sep 27, 2021 09/27/21

Movies and TV shows require multiple takes per scene and may have a substantial amount of archival footage as a result. Here, we use Azure Cognitive Services and custom code to develop a multi-model Machine Learning (ML) solution to automatically detect discardable footage to save media companies manual archiving hours and storage space.

Building an Action Detection Scoring Pipeline for Digital Dailies

Samuel Mendenhall September 20, 2021 Sep 20, 2021 09/20/21

Media companies capture footage filmed for the entire day in what's known as ‘digital dailies’. When talking about terabytes and petabytes of content, storage costs can be a factor. Lets explore Machine Learning approaches to identify which content can be archived or discarded which will save on those storage costs.

Building A Clinical Data Drift Monitoring System With Azure DevOps, Azure Databricks, And MLflow

Nile Wilson October 29, 2020 Oct 29, 2020 10/29/20

Hospitals around the world regularly work towards improving the health of their patients as well as ensuring there are enough resources available for patients awaiting care. During these unprecedented times with the COVID-19 pandemic, Intensive Care Units are having to make difficult decisions at a greater frequency to optimize patient health ...

Real-Time Time Series Analysis at Scale for Trending Topics Detection

Omri Mendels January 2, 2019 Jan 2, 2019 01/2/19

Managing big cities and providing citizens public services requires municipalities to have a keen understanding of what citizens care the most about. And, as the cities we live in seek more and more to become the 'smart cities' of tomorrow, this means gathering and analyzing vast amounts of data. Data gathered from social and municipal sources...

Unsupervised driver safety estimation at scale, a collaboration with Pointer Telocation

Omri Mendels July 30, 2018 Jul 30, 2018 07/30/18

Background Connected cars, which have the ability to collect and transfer data via cellular networks, offer us the opportunity to make road transport safer, more organized and more ecologically friendly. To push forward widespread adoption and help the technology fully enter the mainstream, manufacturers and third parties, like insurance ...

Semantic Segmentation of Small Data using Keras on an Azure Deep Learning Virtual Machine

Margaret Meehan July 18, 2018 Jul 18, 2018 07/18/18

Introduction As previously featured on the Developer Blog, golf performance tracking startup Arccos joined forces with Commercial Software Engineering (CSE) developers in March in hopes of unveiling new improvements to their "virtual caddie" this summer. Powered by Microsoft Azure, Arccos' virtual caddie app uses artificial ...

Sequence Intent Classification Using Hierarchical Attention Networks

Olga Liakhovich March 6, 2018 Mar 6, 2018 03/6/18

Introduction In this code story, we will discuss applications of Hierarchical Attention Neural Networks for sequence classification. In particular, we will use our work the domain of malware detection and classification as a sample application. Malware, or malicious software, refers to harmful computer programs such as viruses, ransomware...

Comparing Image-Classification Systems: Custom Vision Service vs. Inception

Clemens Wolff December 5, 2017 Dec 5, 2017 12/5/17

We recently worked with Reverb, an online marketplace for music gear. Musicians and music retailers use Reverb's platform to buy and sell items such as guitars, keyboards, effects pedals, etc. A key element that aids discovery of items listed on the Reverb site is accurate metadata. For example, the manufacturer's official name for the finish ...

Stock Market Predictions with Natural Language Deep Learning

Patty Ryan December 4, 2017 Dec 4, 2017 12/4/17

Overview We recently worked with a financial services partner to develop a model to predict the future stock market performance of public companies in categories where they invest. The goal was to use select text narrative sections from publicly available earnings release documents to predict and alert their analysts to investment ...

Autoscaling Deep Learning Training with Kubernetes

Rita Zhang November 21, 2017 Nov 21, 2017 11/21/17

Background We recently partnered with Litbit, a San Jose-based startup, on a project to autoscale deep learning training. Litbit enables its customers to turn their "Internet of Things” into conscious personas that can learn, think, and do helpful things. In order to accomplish this goal, customers train their AI-empowered personas using ...

CSE Developer Blog

Machine Learning (ML) - CSE Developer Blog