ANS Center Logo
Search
(0)
  • Home
  • Shop
  • Products
    • ANNHUB
    • DLHUB
  • Downloads
    • ANNHUB
    • DLHUB
  • Blog
  • Contact
  • Help
    • ANNHUB
  • Register
  • Login
  • ANNHUB Banner Image
    ANNHUB & DLHUB
    Easily Bringing Artificial Intelligence Into Your Applications.
    SHOP NOW
  • DLHUB Banner Image
    Multiple platforms
    API Supports For Multiple Programming Languages.
    SHOP NOW
Easy to use Graphical interface. No coding required. Design assistance is supported.
Advanced technologies Machine learning and deep learning with multicore and GPU support
Fast deployment Easy to deploy AI solution into real-time applications

Products

  • ANNHUB
    ANNHUB
    Powerful and user-friendly machine learning tool to speedup designing and deploying a neural network in any LabVIEW, LabWindow CVI, C, C++, C#, iOS, and Android applications. ANNHub supports multi-layer feed-forward Neural Network and Bayesian Neural Network. Evidence framework to find an optimal Neural Network structure is also supported in ANNHub.
    Product Information
  • DLHUB
    DLHUB
    Powerful and user-friendly deep learning tool to speedup designing and deploying a deep learning neural network system in any LabVIEW, LabWindow CVI, C, C++, C#, iOS, and Android applications. DLHub supports Convolution Neural Network and Recurrent Neural Network.
    Product Information
  • NNC
    NNC
    Neural Network Controller for nonlinear dynamic control system.
    Product Information

How DLHUB works?

  • Load dataset
  • Configure
  • Train
  • Evaluate
  • Test
  • Export
  • Deploy
Load training dataset into deep learning DLHUB
DLHUB supports different types of data format, including text and image. Designer can load datatset from a file or folder (1), and confirm if the data is image type (2). Image shape then can be pre-populated or entered manually.
Configure deep learning neural network model in DLHUB
Designer can easily construct any deep learning neural network model by selecting a deep learning node (1), configuring its parameters (2), oganising nodes to form a deep model (3), and verifying constructed deep learning model (4) before proceeding to training procedure.
Train deep learning neural network in DLHUB
In DLHUB, a designer can easily select any training algorithm and specify its parameters (2), configure stopping conditions (2), and start training process (3). DLHUB will detect if the host has GPU to be used for maximising training speed. Training loss and evaluation graphs provide a designer good visual feedback to tweak training parameters in order to achieve the best result.
Evaluate deep learning neural network model with test dataset in DLHUB
Evaluation process is supported in DLHUB to give a designer an indication whether the trained deep learning model meets designed requirements for deployment. In this evaluation page, a designer can load a test set from a file or folder (1) and perform evaluation (2).
Perform prediction/classification with new dataset for trained deep learning model in DLHUB
DLHUB supports testing new dataset inside the software that helps a designer to verify how the trained model works in a deployment environment. A designer can load completely unknown test set from a file or folder (1), select input data to test (2), and visualise the obtained result and predicted class (3).
Export a trained deep learning model to different programming environments.
When a trained deep learning model has been evaluated, tested and verified, it is ready for deployment. Depending on license options purchased from ANS CENTER Shop (1), DLHUB can export trained model to different environments (2).
Deploy trained deep learning model in LabVIEW environment using DLHUB LabVIEW API.
After a trained deep learning model is exported to a model file, it can be loaded into a program developed by a certain programming language. That program will then be compiled to a target executable file to be used in a production. In this example, LabVIEW Application Programming Interface (API) is provided to allow loading a trained deep learning model into a LabVIEW application and performing prediction/classification task. The API will be available in Customer Portal -> Download API menu.

How ANNHUB works?

  • Load dataset
  • Configure
  • Train
  • Evaluate
  • Test
  • Export
  • Deploy
Load dataset into machine learning ANNHUB software.
ANNHUB supports loading dataset in comma-separated values (csv) format. The header, first row, provides dataset' description: input variable names and output variable names. The key words for output variable names are "output, target, and class". Each row contains data for 1 sample.
Configure neural network in machine learning ANNHUB.
ANNHUB simplifies machine learning process. Designer can select training engine, configure a neural network structure, choose pre-processing and post-processing, and apply a certain cost function type for training algorithm.
Train neural network in machine learning ANNHUB.
Designer easily can specify stopping criteria and training parameters for training process. Either early stopping or regularization technique will be used to prevent over-fitting issue during training process. .
Evaluate trained neural network in ANNHUB software.
Evaluation techniques such as confusion matrix, ROC curve, regression curve can be used to evaluate the obtained trained neural network model. The valuation will be performed in both training dataset, validation dataset, and test dataset.
Test trained neural network model with new test set in ANNHUB.
ANNHUB also allows a designer to test a trained model with uknown new test dataset to verfiy how a trained model performs in a simulated production environment before deployment.
Export trained neural network model to different programming environments.
Depending on license options purchased from ANS CENTER Shop, a trained model can be exported in different formats that are supported in different programming environments for deployment.
Deploy trained neural network in LabVIEW environment.
A trained neural network model then can be loaded into any programming language using provided APIs. These APIs are availabe in Customer Portal -> Download API menu. In this example, LabVIEW APIs are used to load a trained neural network model into LabVIEW enviroment to perform a classification task. This LabVIEW application will then be complied into a target executable file for production.

Blog news

Keep in touch with the latest blogs & news.

Image Description
  • |
  • 7/2/2019 1:02:34 AM

How to teach a kid to design a deep learning model that can recognize Marvel superhuman

0
Image Description
  • |
  • 7/1/2019 3:52:26 AM

How to design a time series modeling system and deploy it into Arduino devices

0
Image Description
  • |
  • 6/30/2019 12:56:28 PM

Real-time Vietnamese Speech Recognition

0
Image Description
  • |
  • 6/28/2019 8:49:49 PM

Deep Learning Cifar-10 Classifier Using Residual Neural Network

0
Image Description
  • |
  • 6/28/2019 6:23:21 PM

Fruit Recognition Using Transfer Learning

0
Image Description
  • |
  • 6/28/2019 2:33:36 PM

Convolution Neural Network for MNIST hand writing recognition

0

Latest | tutorial videos

Keep updated with new tutorial videos

Image Description
Play video

Smart auto checkout for supermarkets using DLHUB deep learning in LabVIEW

Real-time LabVIEW Fruit Recognition Using Deep Learning that helps speed-up checking out process in supermarkets and preventing thieves.  

Image Description
Play video

Arduino IRIS Classifier Using Machine Learning

We are going to design a Neural Network classifier to recognize IRIS flowers. This Neural Network classifier is deployed directly into Arduino devices for real-time applications.  

Image Description
Play video

Bayesian Neural Network Time Series Arduino Modelling System

We are going to design a time series modelling system using Bayesian Neural Network. This system will then be deployed directly into Arduino devices for real-time forecasting applications.

Image Description
Play video

Real-time Thought Control System Using Machine Learning in LabVIEW

This project aims to design a Thought Control System that allows severely disabled people to navigate their wheelchairs with just the power of their brain. By using a two-channel wireless EEG system, human brain signal is acquired and then recognized by a machine learning technique.  

Image Description
Play video

Deep Learning LabVIEW: Fruit Detection

This video demonstrates how to use deep learning in LabVIEW to design a real-time fruit detection application that can correctly recognize different types of fruits. More importantly, the expensive NI Vision Development Module is not required in order to develop this native deep learning LabVIEW application.  

Image Description
Play video

Graphical deep learning design for LabVIEW, NXG, C, C++, C#, Android, and iOS

This video demonstrates a new way of designing a deep learning neural network by using friendly user interface software DLHUB. No Python code is required during the deep learning design process. Model verification is supported to check if the constructed deep learning model is valid. Model evaluation and model testing are also supported. The trained deep learning model will be exported directly into LabVIEW, LabVIEW NXG, C, C++, C# environments for real-time deployment.

Image Description
Play video

Machine Learning LabVIEW: Crab Classification

This video provides step by step guide of designing a Bayesian Neural Network that can classify the sex of a crab based on six physical measurements. This trained Neural Network will then be deployed directly into a LabVIEW application for deployment.  

Image Description
Play video

How to design a LabVIEW prediction application using machine learning

In this example, we are going to design a sale prediction system for a game company to estimate the total earning of their new game based on its attributes. Based on historical sale data, we will design a prediction system using machine learning. Most importantly, ANNHUB allows exported this prediction system directly to LabVIEW for actual production

ANS Center provides elegent solutions to simplify machine learning and deep learning design and deployment.

We support mutiple hardware platforms and programming languages, including LabVIEW, LabVIEW NXG, LabWindow CVI, C/C++/C#, and Arduino.

Latest Posts

How to teach a kid to design a deep learning model that can recognize Marvel superhuman

7/2/2019

How to design a time series modeling system and deploy it into Arduino devices

7/1/2019

Real-time Vietnamese Speech Recognition

6/30/2019

Useful Links

  • Home

  • Blogs

  • Tutorials

  • Contact Us

Helps

  • ANNHUB Help

  • ANNHUB API

  • DLHUB API

2018 © All Rights Reserved.
  • Privacy Policy
  • |
  • Terms of Use