ANSTS is a powerful, intuitive platform that allows engineers, educators, and system integrators to design, train, and deploy computer vision models — without writing a single line of code.
ODHUB simplifies every step of the computer vision training workflow. Whether you're working with object detection, classification, segmentation, anomaly detection, ANSTS lets you focus on outcomes — not algorithms.
No-Code, Engineer-Friendly Workflow
Build, train, and evaluate CV models using real-world data without coding or AI expertise, cutting down reliance on external AI consultants or complex toolchains.
Secure, On-Premise Training
All training runs locally, ensuring full control over your data and model IP — ideal for organizations with strict privacy, security, or compliance requirements.
Deployment-Ready Output
Models trained in ANSTS can be exported and deployed directly to LabVIEW environment or Intel/NVIDIA GPU servers
How It Works
ANSTS will automatically validate the correctness of data, including image and label files. It also provides a built-in labeling tool, allowing users to perform labeling tasks and convert labels if necessary.
Key Features
Flexible Model Types for Any Task
Build AI models tailored to your needs. ANSTS supports multiple computer vision tasks, so you can handle object detection, classification, or anomaly detection with ease.
Object Detection
Identify and locate multiple objects within an image or video
Oriented Object Detection
Detect objects with precise orientation, ideal for tasks requiring angle-aware localization
Classification
Categorize images into predefined classes for automated sorting or recognition
Abnormal Detection
Detect irregularities or unusual patterns for quality control or monitoring tasks
Segmentation
Detect objects and precisely identify their shape and boundaries
Built-in labelling tool
ANSTS provides an integrated labeling environment, enabling precise and efficient annotation of your datasets. Support for multiple data formats ensures compatibility with your workflow
.png)
Drag-and-Drop Labelling
Quickly annotate objects with bounding boxes. Label multiple object types in one image
Auto Object Detection
Automatically identify and draw bounding boxes for all detectable objects; users only need to assign class names.
Reuse Existing Labels
Allow importing YOLO, Pascal VOC, and COCO label formats and resuse in ANSTS
Image Augmentation
Enrich datasets with tiling, rotation, flipping, brightness adjustment, and more.
Auto-Labeling
Accelerate labeling by automatically generating annotations based on pre-trained models.
Object Class Management
Easily define, edit, and organize classes for consistent labeling
Efficient Model Training
ANSTS provides an integrated labeling environment, enabling precise and efficient annotation of your datasets. Support for multiple data formats ensures compatibility with your workflow
GPU-Accelerated Training
Fast and efficient training using local NVIDIA GPUs.
Simplified Parameter Setup
Only essential training parameters need to be tweaked, making setup easy for users of all levels.
Performance Monitoring
Training halts automatically if no improvements occur, with real-time training and validation graphs for informed decisions
Model Evaluation & Testing
ANSTS provides an integrated evaluation environment to assess model performance before deployment. Test models on images or live video to ensure accuracy and reliability.
Built-in Evaluation Metrics
Monitor accuracy, precision, and other key metrics to validate model performance.
Real-Time Performance Testing
Observe object detection results instantly for immediate feedback
Flexible Data Input:
Evaluate models using any image dataset or live video from connected cameras
Flexible Model Deployment
ANSTS enables seamless deployment of trained models with optimized formats for different environments, ensuring high performance and broad compatibility.
LabVIEW Deployment
Export models as a .zip package containing architecture, weights, and biases. Deploy natively in LabVIEW using ANSDL for seamless integration into automation and monitoring workflows.
Intel/NVIDIA GPUs Deployment
Run models on Intel or NVIDIA GPUs using ANSVIS, enabling real-time inference, advanced AI features, system integration, and scalable AI vision applications.
Trusted by the global leading companies
