All posts by Daria

Nick @ SPRINT ROBOTICS Robotics Showcasing and Demonstration Events

05/24/2018education articles#Aerial Vehicles, #AI, #California, #Crawler, #Ground Vehicles, #Multicopter, #Robotics Showcasing and Demonstration Events, #Software for self-navigating drones, #SPRINT ROBOTICS, #Underwater VehiclesDaria

It is part of our company’s culture to take part in IT Conferences around the world. We go for many reasons, as speakers, to learn about the latest tech and trends, to meet people with the same tech passion, and, last, but not least, to meet companies with which to collaborate.

We attended two conferences in May, the first was the two-day SPRINT ROBOTICS 2018 Robotics Showcasing and Demonstration Events in Richmond and Pittsburg, California. (We’ll tell you about the other next week.)

SPRINT ROBOTICS describes the event this way.

“Using robotics in the domain of technical inspection and maintenance of capital intensive infrastructure is of vital importance because of the urgency to minimize the impact on safety and the environment. Developments in this domain are moving very fast, and we invite you to come and see some of the solutions that the market has to offer!.”

NTR Lab was one of the participants, along with

In the short video below you can see the introductions of different companies and their robotics.

Thanks to Fiori Films for the video.

Watch our CEO Nick Mikhailovsky and our drone at 1:40.

Source: https://www.youtube.com/watch?v=9K4MtyuhIl8

Continue reading Nick @ SPRINT ROBOTICS Robotics Showcasing and Demonstration Events →

Why Do Neural Networks Need An Activation Function?

04/26/2018education articles#A. Pinkus, #Activation Function, #Approximation Theory, #Computer Vision, #Continuous Functions, #Family of Functions, #Kolmogorov - Arnold, #Neural Networks, #Polytope, #Ridge Functions, #Riemann IntegrableDaria

by Computer Vision Department of NTRLab

Suppose we are given a set of distinct points P = {(x_i, y_i) ∈ ℝ^m ×ℝ}_i=1,…,n which we regard as a set of test samples x_i ∈ ℝ^m with known answers y_i ∈ ℝ. To avoid non-compactness we may assume that P lie in some compact K, for example, K may be some polytope. Does there exist some continuous function in space of all C(K) continuous functions on K such that its graph is a good approximation for our set P in some sense?

From the approximation theory point of view, a neural network is a family of functions {F_θ, θ ∈ Θ} of some functional class. Each special neural network defines each own family of functions. Some of them might be equivalent in some sense. If we restrict ourselves to only MLP according to the above problem with only one intermediate layer consisting of N elements then the corresponding family of functions will be

FORMULA NN 12

Continue reading Why Do Neural Networks Need An Activation Function? →

Demystifying Product/Market Fit

04/11/2018education articles#12 Things about PMF, #CAC, #client segment, #COGS, #conversion instruments, #Cost of Goods Sold, #Customer Acquisition Cost, #entrepreneur, #Eric Ries, #Expected LTV, #Four Steps to the Epiphany, #magic, #Marc Andreessen, #money, #Pmarca, #PMF, #product/market fit, #Profit, #Sean Ellis, #startup, #Startup Marketing, #Startup's Pyramid, #Steve Blank, #success, #the only thing that matters, #The Pmarca Guide to Startups, #traffic source, #Tren Griffin, #user acquisition channel, #you know it when you see itDaria

by Nick Mikhailovsky, CEO and Co-Founder of NTRLab.

Marc Andreessen said, “…the life of any startup can be divided into two parts – before product/market fit and after product/market fit.” It’s a kind of magic thing, without which a startup stands little chance of success. That said, how do you define it beyond “you know it when you see it?”

Continue reading Demystifying Product/Market Fit →

The Main Elements Of The Navigation Stack

04/05/2018education articles#2D, #AI, #AMCL, #ARCore, #Base frame, #Build a map, #Cloud of points, #Drawing up a map, #gmapping, #Hector_mapping, #Kinect, #lidar, #Linear and angular velocity, #Location, #Map, #Map_server, #Move, #Move_base, #Navigation Stack, #Odometer, #Odometry source, #Overcoming the obstacles, #Position, #Robot, #Robot configurations, #Robot navigation, #Robot Operating System, #ROS, #ROS Navigation, #Route, #Sensor sources, #Simultaneous Localization And Mapping, #Space, #Tango, #tf transforms, #Visual odometry, #WheelDaria

By Leonid Bulyga, Robot Navigation System Developer at NTRLab.

To successfully function, a robot needs to know its place in space, in other words, its exact location on the map.

The Navigation Stack, which includes SLAM, allows the robot to build a map, determine its position on it and move around relying on its “feelings.”

Let us consider in more detail what each element of the navigation stack is responsible for.

for blog post

Continue reading The Main Elements Of The Navigation Stack →

Coding Won’t Cut It: How to increase your deep learning skills

03/22/2018education articles#Adam, #AI, #Algorithms, #ArtificialIntelligence, #BayesianApproach, #C, #Calculus, #CNN, #Coding, #ComputerScience, #ConvolutionalNeuralNetworks, #DeepLearning, #DeepLearningExpert, #DL, #DLFrameworks, #ExpertSystems, #FuzzyLogic, #GeneralArtificialIntelligence, #GradientDescent, #healthcare, #HighDimensionalCases, #implementation, #LinearAlgebra, #Lua, #MathSkills, #ML, #ntrlab, #Onotation, #OptimizationAlgorithms, #ProbabilisticFrameworks, #Python, #PyTorch, #ReinforcementLearning, #RMSprop, #robotics, #SelfDrivingCars, #skills, #SoughtAfter, #StatisticsAndProbabilityTheory, #TechCompany, #TensorFlow, #TheVisualCortexDaria

One question always pops up during every Artificial Intelligence (AI) related discussion: How do I become Deep Learning Expert? What do I need to learn?

AI related positions are in high demand and extremely well paid. Everyone wants to become an AI expert and every tech company wants AI experts on their team.

Deep Learning Blog post 2
Continue reading Coding Won’t Cut It: How to increase your deep learning skills →

Info From an Indoor Drone Navigation Developer

03/01/2018education articles#2D/3D suspension, #AerialVehicle, #AI, #aircraft, #ArtificialIntelligence, #ArtificialNeuralNetworks, #BuildAMap, #camera #computer, #compass, #ConvolutionalNeuralNetworks, #coordinates, #GPS, #HectorSLAM, #lidar, #light, #LightSource, #MobileRobot, #NeuralNetworks, #ntrlab, #odometry, #positioning, #RecurrentNeuralNetworks, #RNNs, #robotics, #SimultaneousLocalizationAndMapping, #UnknownEnvironment, #UnknownLocation, #UnmannedAerialVehicle, #VisualOdometry, challenge, drone, indoor, SLAM, UAVDaria

As you may already know, one of the projects that our developers worked on (the expert AI team at NTR Lab) was a project for drones that can work indoors. NTR Lab created unique, embedded software for UAV autonomous indoor navigation.

The original object of the SLAM method was quite narrow. But more and more, as this technology becomes popular and more applicable on practice, we are moving swiftly towards a future we watched in movies.

Continue reading Info From an Indoor Drone Navigation Developer →