High-precision AI-Powered Automated Inspection
Preferred NetworksVisual Inspection

Preferred Networks Visual Inspection has been adopted by over 250 customers

Preferred Networks Visual Inspection is a software that uses proprietary deep learning models (AI) to achieve high-precision automation of visual inspection tasks that were previously difficult to implement.

By achieving high-precision inspections even with limited training data through simplified annotation methods, this software enables rapid and cost-effective inspection system deployment.

This product has proven effective across diverse industries including automotive, electronics, steel production, food processing, semiconductors, and construction, with adoption by over 250 customers since its launch.

See Use Cases

Examples of use cases

Powerful Features Driving Inspection Automation

PFN's proprietary deep learning model

PFN's proprietary excellent deep learning model

Preferred Networks (PFN)'s proprietary excellent deep learning model enables high-precision inspection automation even with minimal training data.

Requires minimal training data

High precision with a small amount of training data

Unlike conventional methods requiring tens of thousands or even hundreds of thousands of training data, our system can train models using as littel as 100 good samples and 20 defective samples.

No detailed annotation required

No detailed annotation required

You only need to divide the prepared product images into "good" and "bad(defective)" groups. The AI will automatically determine the defective areas to focus on.

Defect visualization

Visualize defective parts

By visualizing defective parts through heatmaps, we improve the explainability of the AI's output results.

Intuitive GUI-based training and management interface

Intuitive operation is possible with a GUI tool

Our intuitive GUI-based training tool enables seamless management from image registration to model training and accuracy performance comparison, allowing you to easily build high-precision models.

Flexible and rapid system deployment

Providing an inference library enables flexible and rapid system startup

We provide the inference library necessary for building inspection systems. This seamless integration from training to system deployment enables rapid implementation of inspection automation.

Learn More

Customer Case Studies

Preferred Networks Visual Inspection has been adopted by over 250 customers since its launch and is supporting inspection automation in various industries such as automobiles, electronic devices, steel, food, semiconductors, and construction.

We'd like to introduce case studies of this product.

See Case Studies

TOREFEEDER® of Toray Precision Co., Ltd.

Demo Videos of Inspection Automation

We introduce examples of inspection automation using this product (target work: carpet and metal parts) in a video.

Detection targetCarpet
Type of defectStains and other dirt
Number of training images (good products)80
Number of training images (defective products)40
Detection targetMetal parts
Type of defectScratches
Number of training images (good products)50
Number of training images (defective products)50

Please feel free to contact us.

Contact

Return to top

News

2025/06/03

On 6/4 (Wed) at 11:20, we will broadcast a webinar on PFN's AI visual inspection software Preferred Networks Visual Inspection.

"Innovation of the textile industry supply chain through the introduction of AI visual inspection / Preferred Networks' AI visual inspection with over 200 customers adopted"

It will be broadcast on the video distribution platform Aperza TV. Please register and watch from the link below.
https://tv.aperza.com/event/301?webinar=2010

2024/07/22

Selected as a 2024 NEXT AI TREND by the AI portal media "AIsmiley"

Our Preferred Networks visual inspection software, Preferred Networks Visual Inspection, was selected in the image recognition and image analysis category as an "AI product expected to see a surge in future demand" in the 2024 NEXT AI TREND by AIsmiley (Shibuya-ku, Tokyo, CEO: Koji Itaba), which operates the AI portal media "AIsmiley".
https://aismiley.co.jp/aismiley-ai-products-award-2024-summer/

Award content
"Preferred Networks Visual Inspection by Preferred Networks, Inc. is an AI software that achieves high-precision visual inspection with a small amount of training data. It eliminates the need to fill in defective areas and improves over-detection with its proprietary algorithm. This was highly evaluated for its ability to perform inspections more efficiently and accurately than conventional manual inspections, making a significant contribution to corporate quality control."

2024/01/25

AI Visual Inspection Software Preferred Networks Visual Inspection Webinar Broadcast

What you should know before introducing it! Common problems and how to choose AI visual inspection
ーーKey points of the AI visual inspection system adopted by over 150 customers

A webinar on PFN's AI visual inspection software, Preferred Networks Visual Inspection, which has been adopted by over 150 customers, will be held on the video distribution platform Aperza TV from 1/31 at 10:00. Please register and watch from the link below.
https://tv.aperza.com/event/208?webinar=1317

2023/05/12

AI Visual Inspection Trial Campaign has started

2022/09/27

We will exhibit at Smart Factory Japan 2022 (10/19-21)
PFN to Exhibit Visual Inspection Software at Smart Factory Japan 2022

2021/08/03

The case study of Toray Precision Co., Ltd. has been published
Customer Case Study - Toray Precision Co., Ltd.

2020/12/15

We gave a lecture including the latest case studies at the Quality Forum 2020
Quality Forum 2020 Lecture MaterialsCase Study Excerpts

2019/10/31

The exhibition at CEATEC 2019 was introduced on Robot Start
What did PFN exhibit at CEATEC 2019? Its capabilities are not limited to the tidying-up robot

2019/10/28

The exhibition at CEATEC 2019 was introduced in the Nikkan Kogyo Shimbun
Close-up / Expanding applications of image recognition: AI ventures propose new value

2019/10/15

We exhibited at CEATEC 2019
CEATEC 2019 Official Website

2019/01/09

Featured in the January 2019 issue of Nikkei Monozukuri
Book introduction page

2018/12/12

We held a demo exhibition at SEMICON Japan 2018
SEMICON Japan 2018 - Preferred Networks