Neurocle News

[Newsletter Vol. 5] Captivating the World! Neurocle’s AI Vision Recognized on the Global Stage

Hello, this is Neurocle’s May newsletter, greeting you again after some time!

Over the past two months, the Neurocle team has been extremely busy preparing for three domestic and international exhibitions. In particular, May was a highly rewarding time, as we not only had a lot to prepare but also achieved meaningful results.

Among them, the most exciting news is that Neurocle has won the ‘Innovators Awards’ for five consecutive years with its patch-based algorithm technology. In addition, at Computex, Asia’s largest IT exhibition, we successfully demonstrated that we can provide an edge-based AI vision inspection solution through collaboration with our hardware partner, Portwell. Let’s take a closer look at what technologies have brought Neurocle global recognition.


Proven Technology with 5 Consecutive Wins at the Innovators Awards

This year again brought great news for Neurocle. We achieved the remarkable milestone of winning the ‘Innovators Awards,’ hosted by Vision Systems Design, the world’s leading machine vision media, for five consecutive years. This award is especially meaningful as it recognizes both the innovation and practicality of Neurocle’s recently developed patch-based high-resolution image inspection technology in industrial environments.

Recently, inspection processes utilizing high-resolution images of 4K or higher have become essential in industries such as semiconductors, precision electronic components, and batteries. However, there are still limitations in processing such large-scale data.

To overcome this, Neurocle has continuously advanced its patch-based deep learning algorithms. For example, the Patch Classification model divides images into patches to accurately detect even fine defects, while the Patch Mode of the Segmentation model can distinguish ultra-small defects at the scale of tens of micrometers (μm), enabling the identification of electrode coating defects and surface scratches on films. In addition, by applying patch technology to a generative defect model (GAN), it has become possible to generate high-quality data similar to real processes. As a result, fast and stable AI model deployment is achievable even for new production lines.

With this capability to reliably detect fine defects in high-resolution images of hundreds of megabytes, Neurocle’s technology is now establishing itself as a solution that addresses real challenges in industrial environments.


Even on Edge Devices? AI Vision Inspection Solution Recognized in Taiwan, the Heart of AI

Neurocle showcased an edge-optimized deep learning vision solution at COMPUTEX 2025, Asia’s largest IT exhibition, in collaboration with Portwell, a Taiwan-based industrial computing company. By deploying Neurocle’s deep learning models on Portwell’s NVIDIA Jetson-based hardware (PJAI series), we enabled fast and accurate defect detection even in constrained computing environments.

In edge environments, limited computational resources have often made it difficult to operate high-performance deep learning models. Neurocle overcame this limitation through its auto deep learning algorithm, which automatically optimizes model structures and hyperparameters, enabling a lightweight AI inspection system that performs high-precision inspection without the need for a separate server.

One of the most notable demonstrations at COMPUTEX was the real-time detection of defects from actual industrial processes. Neurocle’s model achieved approximately 2.5 times faster inference speed compared to similar existing specifications, proving not only that it can run on edge devices, but also that it delivers practical levels of accuracy and speed for real industrial applications.

The technology demonstrated at this exhibition showed the potential to achieve server-level performance even in edge environments and served as an opportunity to highlight both the starting point of AI vision solutions tailored to global industrial needs and the scalability of Neurocle’s technology on the global stage.


Through these achievements, Neurocle has once again gained recognition for both its technological excellence and practicality at home and abroad, further strengthening its position in the global AI vision market. Neurocle will continue its journey to solve real-world industrial challenges and develop AI vision solutions that are easy for anyone to use.

Stay tuned for the next newsletter to see what updates Neurocle will bring!