Here at Thread Check Co. we believe that even the most tactile of markets can benefit from the impacts of digital applications. Our mission is to help bring fabric classification and defect detection into the modern age.

The technology behind our innovations...

All of our products at Thread Check Co. utilize the power of artificial intelligence (AI). Machine learning is a type of AI that involves learning from data inputted, and using that for various applications. It involved using trained algorithms that observe similarities in patterns and markings. Specifically, an unsupervised neural network is used for Hot-Shatz. A neural network recognizes the relationships between data with methods that mimic the way that neurons operate in the brain. It involves an input layer where items are assigned, a second layer where calculations are made, and an output layer where conclusions are made. For our neural networks at Thread Check Co. we inputted microscope images of different fabrics and threads, making up the first layer. The second layer is the layer in which we put the program through various forms of training once all the images are inserted. The last layer, the output layer is the determination or classification of the type of fabric or thread that the user inputted into our app.

Design and methodology

Here at Thread Check Co., we specifically use the machine learning open source library Tensorflow.js. This program utilizes the method of transfer learning, in which a pre-trained model is reused and recategorized in order to fulfill a new task. Using image recognition and machine learning, we trained our apps to properly assign fabric names or fabric qualities to visual inputs. This has given the base neural network a plethora of new abilities. To learn more about some of the basics types of AI explore the drop down menu below!