top of page
luke-chesser-CxBx_J3yp9g-unsplash.jpg

DLHUB

Deep Learning Studio

Design, train, and validate custom deep learning models with a few simple clicks. No programming skills are required.

How It Works

DLHUB supports different types of datasets, including text files, image files, and image folders. When a dataset is loaded into DLHUB, the data normalization procedure is automatically applied.

DLHUB Load dataset

Product Features

No coding

DLHUB is specifically designed for engineers who do not have a profound knowledge of programming skills to design a proper deep-learning model in only a few simple clicks. It provides a graphical user interface (GUI) that makes it easy to select the appropriate components, configure the model, and train it.

  • A library of pre-trained models that can be used as a starting point for transfer learning. This can save engineers a lot of time and effort in the model development process.

  • A drag-and-drop interface that makes it easy to select and configure the different components of a deep learning model. 

  • A visual training dashboard that allows engineers to monitor the training process and to make adjustments to the model as needed. This can help to ensure that the model is trained properly.

  • Automatically detects and enables NVIDIA GPU for the training process. This can significantly speed up the training process, especially for large models.

Model exporting
DLHUB_RearrangeNodes
DLHUB neural network module
DLHUB_RearrangeNodes
transfer learning parameters

Multiple Deep Learning Modules

DLHUB supports different types of deep learning modules that help users to create and customize their AI model, including:

  • Conv2D

  • Pool2D

  • Dense

  • Flatten

  • Drop

  • ResNet

  • Activation

  • Transfer learning

  • Recurrent nodes

Training parameters

DLHUB supports the training of deep learning models with advanced algorithms, including:

  • AdaDelta: an adaptive learning rate algorithm that is designed to be more robust to noisy gradients than other algorithms.

  • AdaGrad: an adaptive learning rate algorithm that is designed to give more weight to recent gradients.

  • FSAdaGrad: a variant of AdaGrad that is designed to be more efficient.

  • Adam: a popular learning rate algorithm that combines the advantages of AdaGrad and RMSProp.

  • RMSProp: a learning rate algorithm that is designed to be less sensitive to noise than other algorithms.

  • MomentumSGD:  a stochastic gradient descent algorithm that uses momentum to help the model converge faster.

  • SGD: a simple stochastic gradient descent algorithm that is easy to implement

DLHUB_TrainingProcess
DLHUB_Select_TrainingEngine
DLHUB_TestingProcess
DLHUB_Evaluation

Support model testing

DLHUB supports a built-in test interface that allows users to test the model performance before final deployment for production. Users can evaluate the trained model with a new dataset for instant feedback and make necessary adjustments.

Optimized for LabVIEW

DLHUB allows you to export your models in .zip format. The zip file contains all the necessary information, such as the model's architecture, weights, and biases, to deploy in LabVIEW using ANSDL

DLHUB_Deployment_In_LabVIEW
DLHUB_ExportTrainedModel

Resources

System Requirement

Documents

Tutorial Videos

Community

luke-chesser-CxBx_J3yp9g-unsplash.jpg

Are you ready to dive in?

Label data seamlessly, train models efficiently, improve performance iteratively, and deploy applications rapidly—all with ANSCENTER.

bottom of page