A zero-code platform that covers the whole vision-AI workflow: build datasets, train and validate models, quantize and export them, then deploy them to TouchVision edge AI boxes and servers. Teams can ship defect-detection or safety models without a deep-learning specialist.
Overview
Most AI vision projects stall between a working demo and a model that runs reliably on site. TouchVision’s AI Model Training & Inference Platform closes that gap with a zero-code, zero-configuration workflow. Engineers import images, preprocess and augment them, build a dataset, then train and validate one of more than ten mainstream model architectures.
Once the model meets its accuracy target, the platform quantizes and accelerates it for TouchVision’s SOPHGO-based edge AI hardware, exports it and runs inference. A C++ model inference library is provided for embedded integration, and customized editions can support additional model types for specific projects.
Key features
- Supports 10+ mainstream models, from training through deployment and inference
- Zero-code, zero-configuration model training and deployment
- Image preprocessing, data augmentation, dataset creation, training and validation
- Model quantization and acceleration, export, inference and tuning
- C++ model inference library for embedded deployment
- Customized editions support more model types
Specifications
| Item | Specification |
|---|---|
| Workflow | Preprocessing → augmentation → dataset → training → validation → quantization → export → inference |
| Models | 10+ mainstream model architectures; more in customized editions |
| Coding | Zero code, zero configuration |
| Deployment Targets | TouchVision edge AI boxes, AI edge servers, 2U AI servers and AI modules |
| SDK | C++ model inference library |
| Delivery | On-premise; customized editions available |
Specifications are based on manufacturer data and may change. Ask for the latest datasheet for your configuration.
Applications
- Industrial defect detection
- Security and intrusion detection
- Building compliance checks
- Agricultural pest and disease detection
- Biometrics
- Building automation
Related industry solutions: AI Video Analytics & Smart Safety · Industrial Automation & Machine Builders
Customization and ordering
The AI training platform can be customized for OEM projects, and small-batch pilot orders are supported. Request a quote or sample, chat on WhatsApp (+86 135 0532 6001), or read about our OEM/ODM service.
More in this category: Industrial Software & Embedded Control Modules · Related: Edge AI Computers & Video Analytics Boxes · On-Premise LLM & AI Agent Appliances
Frequently asked questions
Do I need deep-learning experience to use the platform?
No. The platform is zero-code and zero-configuration, so engineers can prepare data, train and deploy models without writing training scripts.
How are trained models deployed?
Models are quantized and exported for TouchVision edge AI hardware. A C++ inference library is provided for embedded integration.
Can the platform be customized?
Yes. Customized editions support additional models and project-specific workflows.