DOCSEARCHbyAlgolia
Submit Docs

AIBox Vision Documentation

website·Indexed 31 days ago

https://help.aiboxvision.com

Documentation for AIBox Vision products, including ANSVIS for AI task management and camera integration, ODHUB for computer vision object detection training and evaluation, and DLHUB for deep learning classification tasks.

Tokens
38.7K
Snippets
26
Records
336
Agent score
52%

What's inside AIBox Vision

  1. Overview of ANNHUB Machine Learning Studio

    ANNHUB is a no-code machine learning studio designed for engineers who may lack deep programming or machine learning expertise. It enables users to design, train, and validate machine learning models for classification and prediction applications through a guided, click-through interface with auto-recommended training and validation features.

  2. Overview of ANNHUB for no-code machine learning

    ANNHUB provides a no-code environment to design, train, and validate machine learning models, including tools for dataset preparation and model training.

  3. Overview of DLHUB for no-code deep learning

    DLHUB enables the creation and validation of deep learning models using guided, no-code workflows, including dataset preparation and deep learning modules.

  4. Overview of ODHUB for object detection training

    ODHUB is dedicated to training and validating object detection models for computer vision tasks, providing a labelling tool and workflows for training and testing.

  5. Overview of ANS Labelling Tool

    The ANS Labelling Tool is a desktop application used to create image datasets for computer-vision training. It supports three primary annotation tasks: rectangular bounding boxes (Detection), polygons (Instance Segmentation), and rotated bounding boxes (OBB). The tool uses the 'ANS standard format' (a folder of images with sidecar text label files and an ANS_Class.json definition) but can import and convert datasets from popular formats including YOLO, COCO, Pascal VOC, CreateML, Paligemma, Darknet, and TFRecord.

  6. Overview of ANS License Manager

    The ANS License Manager is a standalone application installed on each target machine to provide decentralized license control. It allows users to activate and deactivate software licenses, view license status and expiration dates, and supports both online and offline activation. Note that ANSVIS software also includes a built-in License Manager as an alternative.

  7. Overview of ANSTS for computer vision model deployment

    ANSTS is a desktop application used to build, train, evaluate, and deploy computer vision models, featuring integrated labelling, training, and testing tools.

  8. Overview of ANS Training Studio (ANSTS) capabilities

    ANS Training Studio (ANSTS) is a desktop application for the end-to-end machine learning workflow of visual AI. It supports the following key capabilities:

    • No-Code Training: Build and train models without programming expertise.
    • Supported Model Types: Object Detection, Classification, Segmentation, and Anomaly Detection.
    • Data Management: Integrated image labeling and support for importing YOLO, Pascal VOC, and JSON annotation formats.
    • Hardware & Security: Local processing using NVIDIA GPU acceleration, ensuring data privacy and offline capability.
    • Analytics: Live visualization of training progress via accuracy charts, loss curves, and confusion matrices.
    • Deployment: Exported models are compatible with ANSVIS (for Edge AI video analytics) and ANSDL (for LabVIEW integration).

  9. Overview of DLHUB Deep Learning Studio

    DLHUB is a no-code deep learning studio designed for engineers to design, train, and validate deep learning models for industrial solutions. It eliminates the need for profound knowledge of deep learning theory or proficiency in Python and C# programming by providing a drag-and-drop interface and built-in model verification features.

  10. Overview of ANSDL for LabVIEW

    ANSDL, developed by ANSCENTER, is a collection of LabVIEW VI toolkits that enable the integration and deployment of AI models and vision functions directly within LabVIEW projects. It allows users to bridge AI models with industrial automation systems without requiring deep AI or programming expertise.

  11. Overview of ODHUB Object Detection Studio

    ODHUB is a no-code object detection studio designed for creating, training, and validating computer vision models. It is used for tasks such as manufacturing defect detection, license plate recognition, people counting, and food detection. The platform supports multiple frameworks to optimize for inference speed and accuracy, including TensorFlow (SSD, Faster RCNN, EfficientDet) and YOLO (YOLO2, YOLO3, YOLO4).

  12. Overview of ANSVIS for video source and AI task configuration

    ANSVIS is used to configure video sources, AI tasks, events, triggers, and deployment workflows.