LT SmartEye Deep Learning Platform

The LT SmartEye deep learning training platform is a high-performance computing platform designed specifically for the development and training of deep learning models. The main functions include object detection, semantic segmentation, classification detection, OCR recognition, and more.

Product Benefits

Effectively deals with defect detection and classification issues in a variety of complex backgrounds

Supports zero-code training and deployment

Seamless integration into LT Vision Builder visual inspection platform

Supports flexible customization and development

Application Scenarios

Metal surface defect detection (fuel injector, rail surface)

Original image of textile surface defect detection (fabric)

Inspection for foreign objects on automotive components

Defect detection on PCB board (short circuit, open circuit)

Original image of solar panels defect detection

Irregular OCR character recognition

Metal surface defect detection (fuel injector, rail surface)

Original image of textile surface defect detection (fabric)

Inspection for foreign objects on automotive components

Defect detection on PCB board (short circuit, open circuit)

Original image of solar panels defect detection

Irregular OCR character recognition

Metal surface defect detection (fuel injector, rail surface)

Original image of textile surface defect detection (fabric)

Inspection for foreign objects on automotive components

Defect detection on PCB board (short circuit, open circuit)

Original image of solar panels defect detection

Irregular OCR character recognition

Metal surface defect detection (fuel injector, rail surface)

Original image of textile surface defect detection (fabric)

Inspection for foreign objects on automotive components

Defect detection on PCB board (short circuit, open circuit)

Original image of solar panels defect detection

Irregular OCR character recognition

Product Function

The platform includes algorithm models for image classification and detection, object detection, and image segmentation.

Supporting unsupervised defect detection algorithms that do not require defective images for training, only normal and defect free images are needed. This can solve the problem of supervised deep learning methods not being able to detect unknown defects. It also has a stronger ability to express image features when compared to traditional methods.