Built libraries and public C headers. 12ms per frame on CPU only, no GPU required. Over 8 years and more than 100 units deployed.
To device manufacturers and system integrators who want to incorporate AI inspection functions into devices that handle images: DeepAge Sensor is provided as a pre-built inference library (Windows/UNIX). You only need to reference a single public C header (extern "C"). It works simply by linking in C/C++ or loading with ctypes in Python. The CPU version requires no GPU, CUDA, or drivers, and operates at 12ms per image (the GPU version operates at 2ms). Training is conducted on our cloud, and only the model file is distributed to the devices, so there is no need to rebuild the device software every time the model is updated. An unsupervised anomaly detection library (.uai) can also be used with the same API structure. We will provide a library and sample model for evaluation, so you can first check if create → load → inference works in your environment. OEM/white label offerings are available. Your company will handle first-level support, while we will handle second-level support.
Inquire About This Product
basic information
【Provided Items】Pre-built inference library (GPU version and CPU version) + public C header (inference.hpp) + model file (.ai/.uai) + complete self-test for operation confirmation 【Operating Requirements】CPU version: No GPU, CUDA, or driver required (12ms per image) / GPU version: NVIDIA sm_86 or later, CUDA 13.2 series (2ms per image) 【OS】Windows / UNIX (Linux x86_64) 【API】Four stages: create → load → inference → close (6 functions). Unsupervised anomaly detection follows the same structure 【Model Update】Only file replacement is needed. No need to rebuild the device software 【Support Responsibilities】Primary: Your company / Secondary: Our company (technical separation and bug response)
Price information
OEM royalty (based on the ratio to the equipment's base price) individual quotation / evaluation library and sample model provided.
Price range
P3
Delivery Time
P3
※Evaluation library is provided immediately.
Applications/Examples of results
- OEM integration of AI inspection functions into products of inspection device manufacturers (over 100 units implemented since the start of provision in 2018) - Supports both retrofitting inspection options to existing models and standard installation in next-generation models - Proven track record of deployment in devices for overseas markets, including China
Detailed information
-
8 years of experience with over 100 devices installed and key specifications.
-
Three learning methods. The unsupervised anomaly detection library has the same API structure.
-
Learning is done in the cloud, while inference is performed solely by the device's CPU. Model updates are done only by replacing files.
-
1. Receipt of evaluation library 2. Operation confirmation with self-test 3. Application integration 4. Model creation in real work (first time free)
-
6 industries and various defects addressed (solder defects/scratches/dents/burrs/chips/foreign objects/missing items/printing defects)
catalog（2）
Download All Catalogs
Company information
We provide services that utilize data. The value generated from data is increasing day by day. While there may be a focus on numbers in sales targets and marketing, the importance of processing large amounts of data with computers to create new value has not been emphasized as much. Technology is evolving every day. We are now able to gather various data in the cloud and process large amounts of data. However, even with an understanding of such technologies and their value, practical application is still not sufficiently realized. We will accurately understand the current issues, sincerely consider whether artificial intelligence and new technologies can solve them, and communicate our findings to the world in the most optimal way.
![Cybersecurity Measures | [KATABAMI VDP] Vulnerability Assessment](https://image.mono.ipros.com/public/product/image/2003003/IPROS17684332589987153228.jpeg?w=280&h=280)