Novetta

NOVETTA NEXUS
Advanced Analytics

Media, Data, and Outbreaks: The UN’s Global Health Crises Task Force

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /

by

The challenges that epidemics and other global health crises present to the international community are ever changing in each new environment that develops potential risk to the adequate containment and eradication of disease outbreaks. In an effort to find new procedures and practices to limit these risks in 2017 the UN Secretary General’s Global Health Crises Task Force created an independent monitoring and accountability body called the Global Preparedness Monitoring Board (GPMB).  In its most recent report produced by the International Federation of the Red Cross and Red Crescent Societies in collaboration with Anthrologica, the GPMB focused on the need...

Read More

ADSynth Blog Series: AWS Pipeline and iOS App Creation

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /

by Matthew Kongsiri, Jack Buttimer

Matthew and Jack interned with Novetta’s Machine Learning Center of Excellence during the summer of 2019. This blog series discusses ADSynth, an app that creates a digital architecture diagram from a photo of a whiteboard sketch.    As introduced in our first blog post, we built a simple iOS application alongside an AWS pipeline to provide an end-to-end solution to diagram creation. AWS was the obvious solution to build a truly scalable system capable of supporting multiple users creating diagrams at once. We were already using GPU enabled EC2 instances during the development of our object detection model, hence most...

Read More


Part 4 of 4 - See a demo at Strata Data Conference Booth 1201!

ADSynth Blog Series: Defining the AWS Architecture Diagram

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /  /

by Matthew Kongsiri, Jack Buttimer

Matthew and Jack interned with Novetta’s Machine Learning Center of Excellence during the summer of 2019. This blog series discusses ADSynth, an app that creates a digital architecture diagram from a photo of a whiteboard sketch.    ADSynth leverages numerous image processing techniques to properly translate a whiteboard sketch into an editable, digital architecture diagram. These techniques are used to ensure that all detected components are connected and formatted correctly. This post describes how we utilized our YOLOv3 object detection model predictions in conjunction with the image of the whiteboard sketch to obtain a clean, presentable graphic ready for insertion...

Read More


Part 3 of 4

ADSynth Blog Series: Object Detection Model and Training Data

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /  /

by Matthew Kongsiri, Jack Buttimer

Matthew and Jack interned with Novetta’s Machine Learning Center of Excellence during the summer of 2019. This blog series discusses ADSynth, an app that creates a digital architecture diagram from a photo of a whiteboard sketch.    We began with the backbone of ADSynth: the object detection model that makes diagram synthesis possible, as well as the data necessary to train the model.  Implementation of YOLOv3 Architecture  Based on the research we conducted on object detection, the architecture we decided to implement was YOLOv3. YOLOv3 is a deep neural network comprising of 106 layers and almost 63 million parameters. The...

Read More


Part 2 of 4

ADSynth Blog Series: Digitizing Whiteboard Diagrams

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /

by Matthew Kongsiri, Jack Buttimer

Matthew and Jack interned with Novetta’s Machine Learning Center of Excellence during the summer of 2019. This blog series discusses ADSynth, an app that creates a digital architecture diagram from a photo of a whiteboard sketch.    This summer, we interned with Novetta’s Machine Learning Center of Excellence (ML COE). As computer science students with a passion for machine learning and problem solving, we were excited to have the opportunity to apply our classroom knowledge in an industry setting. The experience was rewarding, though transferring our theoretical understanding of machine learning to practice presented a steep learning curve (pun intended)....

Read More


Part 1 of 4

Musical Face Match

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /  /  /  /  /  /

by

During my summer internship at Novetta, I developed a prototype that can match a person’s face with the musician whom they most closely resemble, then play a song by that musician. This project was fun, but is also a great demonstration of a highly adaptable concept: face recognition from a live video feed triggering an action based on a specific identification. Development In planning this prototype, I knew I would need to leverage Amazon’s suite of developer applications. My background is in applied mathematics, so I needed to familiarize myself with that platform. Once I learned how the services communicated...

Read More


A Face Recognition Prototype with a Twist

Deploying Advanced Machine Learning Models at Scale

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /  /  /  /  /

by

Novetta is using deep learning to automate advanced analytic pipelines for our customers. These new capabilities identify important pieces of information across large text, image, and video data sets, a challenge faced by many of our Defense and Intelligence Community customers.  We developed Argo, a standardized, Docker-based model training and deployment pipeline, to enable rapid evaluation and deployment of state-of-the-art deep learning models. Argo enables models to be deployed as endpoints and, when used with Amazon SageMaker, it forms the foundation for us to deploy, scale, and manage our deep learning models. Simplifying infrastructure setup through Docker and AWS allows...

Read More

Ebola in the Democratic Republic of Congo: Novetta Mission Analytics

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /

by Rhys O'Neill, Jonas Eno-Van Fleet

The ongoing Ebola outbreak in eastern DRC has caused 1,522 deaths, making it the second largest in history. Understanding the causes and expressions of deep-seated local resistance to international response efforts is essential to shortening the outbreak and saving lives. Novetta is providing hands-on analysis via subject matter experts to our partners in the Ebola response community through the classification and analysis of sentiment and voice share in hyper-local and regional media content originating in the North Kivu Ebola hotspot. Novetta’s analysts identify distinct narratives and vectors of propagation from the voices of citizens, civil society stakeholders, Ebola response leadership,...

Read More


How data analysis is helping mitigate the second largest outbreak in history.

PIPPIN: A Public IP Inspector Capability for AWS

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /  /  /  /

by

‘White hat’ penetration testing, which allows for better understanding and mitigation of system vulnerabilities, starts with full insight into your organization’s publicly accessible systems. Across multiple accounts, services, and regions, there is no standard report available in AWS to quickly and efficiently provide this information.  We created a public IP inspector, PIPPIN, to address this challenge using the AWS Application Programming Interface (API). This post walks through our process of discovering and cataloging an organization’s publicly accessible systems and our results with PIPPIN.    Defining your Public IP Landscape You first need to collect AWS Account IDs, a twelve digit...

Read More

20% Accuracy Bump in Text Classification with ME-ULMFiT

Category: Advanced Analytics / Novetta Nexus Tag(s):  /  /  /  /  /  /

by Matt Teschke, Brian Sacash, Ethan Gumabay

Introduction In a previous post, we discussed an enhanced method of passing additional data to the ULMFiT1 algorithm for unstructured text classification, which improved accuracy by up to 9%. In follow-on experiments, we have demonstrated that this approach – in which we prepend metadata to the text to provide more information to the model – increased accuracy by up to 20%, in relative terms. This improvement required just a few lines of Python, such that the performance boost comes at essentially no cost. This blog post explains the findings of this more-detailed evaluation, where we have compared how the following...

Read More