Core competence

Service mode

AIWin guides customers to use AI,combine deep learning with Domain Know-how,collect data,pre-process data,transform and analyze data, establish AI training and verification model, provide complete AI solutions, and assist customers to import AI positive cycle.

1. Data collection

Mark the large amount of information needed to train AI.

2. Data conversion analysis

Determine the learning data and adjust it so that the machine learns to recognize text, pictures, images, etc.

3. AI model establishment

Establish a variety of AI models and verify the learning results.

4.System Integration

Software design, function development, AI application and system integration.

AI service

AOI flow chart

Introduction of AOI application in products

AI-AOI equipment for industrial functional textile setting

AIWin has developed a fabric AI-AOI technology for the textile industry of Taiwan. It can help user building AI models quickly by AIWin AI Algorithms and provides fast and accurate inspection quality to reduce human fatigue effectively.

  • AIWin Deep Learning and Computer Vision Technology
    Detection Rate: more than 95%

  • Max. Speed: 25m/min.
  • lAIWin Special Mechanism

  • lDetect Defect Type: more than 6 types

  • Optical Resolution: 0.014mm/pixels

 

 

  • The recognition technology is the exclusive deep learning technology of AIWin, and the recognition rate is over 98.5%.

  • The rate of missed detection is less than 0.1% (manual missed detection rate is 1-3%), and the error-detection rate is less than 10%

  • The recognition speed can reach 3,600 pieces / hour.

  • Exclusive turning and in and out mechanism.

  • It can effectively identify more than 28 kinds of golf-related defects and logo defects.

  • The optical accuracy is 0.06mm/pixels.

Based on the shoelace production line, an optical image acquisition system is set up, and the deep learning AI image recognition technology is integrated through the camera to detect the defects on the shoelace surface.

Technical focus

1. One-shot Face Recognition technology is adopted for face recognition. After a single photo is saved, the face can be recognized.

2. Face recognition rate is over 99.80% in public database LFW and over 98% in Megaface.

3. Feature: no need to save face image. It can eliminate privacy concerns.

4. Feature comparison: it can compare a face within 0.03 seconds and a complete 720P image within 0.1 seconds.

5. The API can call across different OS.

6. Nvidia GPU acceleration technology is adopted.

IVS flow chart

Introduction of IVS application in products

The use of license plate recognition can make the parking lot fully automatic and intelligent, which can not only reduce the labor cost but also reduce the trouble of drivers (no need to collect magnetic snap manually). In addition, combined with software, it can achieve the control of data and black-and-white lists of users.

Human pose estimation is to find out the skeleton position through deep learning and solve the behavior, posture and other problems through analysis. The application includes crowd monitoring, abnormal behavior monitoring, sleep posture monitoring, fall monitoring, athlete movement analysis, etc.

The core function of face recognition technology is identity recognition with the image, AI artificial intelligence combined with face recognition algorithm to calculate face features and analyze gender, age group and emotion. Related applications include blacklist detection, VIP identification, access control, unmanned stores, and customer analysis in the retail industry.

AI image recognition technology is used to solve the safety management problem of the production environment and reduce labor costs. The application includes access control, staff safety equipment detection, abnormal situation warning of the monitoring field, warning of dangerous behavior, etc.