Big News｜Aqrose Technology’s "Artificial Intelligent Defect Generator - AIDG" is officially released!
AIDG (Artificial Intelligent Defect Generator) is an AI-based intelligent defect generator that generates defects of different locations and shapes through simulation models to address the problem of insufficient support for model training due to the lack of defect data in the industrial production process. It can be applied to various application scenarios and is highly synergistic with the industrial AI vision software - AIDI to help users efficiently complete model iteration and help the industry to upgrade intelligently.
In the process of industrial production, the following situations often occur:
⭕ With high product yield rate, defective products are difficult to present;
⭕ Product models change and it is difficult to collect enough defective products in a short period of time;
Faced with this situation, engineers would usually use PS tools, or manually create defects on non-defective products, or wait on the production line, etc. to obtain enough defect maps. However, the above approaches would usually lead to the following problems, which would increase the burden on enterprises and factories:
It is difficult to ensure the authenticity of defects by manually creating defects;
⭕ Longer and less efficient project cycle time;
⭕ Higher project costs;
Since its establishment, Aqrose Technology is committed to become the most industry-savvy AI algorithm and product team and bring AI to millions of factories. To address the above problems, Aqrose Technology has developed an AI technology-based intelligent defect generator-AIDG to assist the AI landing.
AIDG employs AI technology to generate a large number of defect maps with only 3-5 defect maps and 2-step operations, making the process of obtaining defect maps more intelligent and efficient.
AIDG functional features are as follows:
● High applicability for different products and different defects
● Strong flexibility for independent selection of defect location and shape
● Easy to use, with 2 steps to generate defect map intelligently
● Realistic and valid, the results can be directly used for model training
● Highly synergistic, results can be imported into AIDI without secondary labeling