NeuroCheck introduces new Deep Learning Toolkit 4

Jul 22, 2026 | NeuroCheck Company

Feature Highlights of the NcDL-Toolkit: WebApp, Labeling and Validiation, Training Analysis and Model Evaluation, Role- und Project-Based Collaboration, On-Premise Operation, Seamless Integration into NeuroCheck
With the NeuroCheck Deep Learning Toolkit 4 (NcDLToolkit 4), NeuroCheck is launching a new central platform for training, managing, and deploying modern deep learning models for production use within NeuroCheck software. As an on-premise solution, all images, training data, and AI models remain entirely within the user’s IT infrastructure. This ensures companies retain full control over their data at all times – without the cloud, without external dependencies, and with maximum data security.
Screenshot Project Management

Browser-Based Web Application

The toolkit supports the efficient development of AI-based machine vision solutions via an intuitive, browser-based web interface. It covers key deep learning techniques used in industrial image processing, including classification, segmentation, object detection, and anomaly detection. This enables the implementation of a wide range of inspection and automation tasks that require the reliable detection of both clearly defined inspection features and complex, previously unknown defect patterns.

Screenshot Labeling of Training Data

Role- and Project-Based Collaboration

Integrated project and user management enables teams to collaborate in a structured manner on the development and management of training projects. Role-based permissions ensure clear responsibilities, while features for image annotation and augmentation, training, validation, and model evaluation ensure an end-to-end, traceable AI workflow.

Screenshot Training Analysis

Training Analysis and Model Evaluation

During training, users gain real-time insights into their models’ performance. Live metrics, heatmaps, confusion matrices, and analyses of model evolution across successive training runs provide well-founded information regarding quality and stability. This enables users to identify optimization opportunities early and refine their models with confidence.

Screenshot ONNX Export

Seamless Integration into NeuroCheck

Deployment-ready models can be exported as ONNX-compatible neural networks and seamlessly integrated into NeuroCheck via the AI-Xtension. Combined with proven classical machine vision methods, for example for image preprocessing or positioning, AI inference can be used selectively where it delivers the greatest value. Together, NeuroCheck 6.5 and NeuroCheck Deep Learning Toolkit 4 provide a powerful foundation for robust, automated, and scalable machine vision solutions for production environments.

Highlights at a Glance

  • Full data sovereignty through on-premise operation
  • Modern, browser-based training platform
  • End-to-end deep learning workflow
  • Efficient teamwork through role- and project-based collaboration
  • Integrated tools for labeling, augmentation and validation
  • Transparent training analysis with meaningful visualizations
  • Standardized ONNX export
  • Seamless integration into NeuroCheck via the AI-Xtension

Would you like to learn more or see the NeuroCheck Deep Learning Toolkit in action? Our sales team will be happy to assist you at sales@neurocheck.com.