August 6, 2020
Adacotech won IVS LaunchPad, a gateway to be successful startup
Adacotech Co., Ltd.(Head office: Chiyoda-ku, Tokyo, CEO: Ryota Kawamura, hereinafter referred to as “Adacotech”) is pleased to announce that it has won the final selection of the pitch event “IVS LaunchPad” held on July 31 (Friday), which is the gateway to success for startups.
Adacotech is a technology company that provides software that enables more efficient anomaly detection than ever before, based on image and video analysis technology using the HLAC(* 1) feature extraction method developed by the National Institute of Advanced Industrial Science and Technology .
This technology is capable of highly accurate analysis with a small amount of data, and is a model that “detects deviations from normal” rather than detecting what it learns as abnormal. Therefore, it is possible to detect almost 100% of abnormalities, including cases that have no precedent.
Another advantage is that it does not require arithmetic operations, so the burden of computational processing is small, and even general-purpose PCs such as notebook PCs can be used for detailed order processing and operation.
At present, we have four types of abnormality detection systems that match the input data of “video,” “still image,” “sound,” and “sensor data,” and we provide them in appropriate combinations according to the application and target to solve the problems of our customers.
As for our past achievements, we have introduced image analysis solutions for inspection and inspection to the actual production lines of major automotive parts manufacturers and other manufacturing companies.In collaboration with Mitsui E&S Machinery Co., Ltd. we have commercialized a technology for non-destructive inspection of tunnel deterioration using radar exploration. In addition, we have provided Central Security Services Co., Ltd. with video analysis software that automates the monitoring of unauthorized entry into large-scale commercial facilities and railroad infrastructure.
IVS LaunchPad is one of the programs of the 「Infinity Ventures Summit 2020」one of the largest startup conferences in Japan run by Infinity Ventures Summit, and is a pitch event where challengers who will create the future gather.
The 14 companies that won out of the largest number of applicants ever took part in the final selection held on July 31st (Friday).
As a solution to the serious shortage of human resources at manufacturing sites, Adacotech
made a presentation on the automation of inspection and testing in manufacturing through image and video analysis using the “HLAC” feature extraction method. The solution using this technology can build an extremely accurate learning model with a small amount of data compared to conventional AI and deep learning based inspection and testing, and can build a model that can detect abnormalities using only normal data without requiring abnormal data such as defective products. In addition, the calculation process and results can be clearly shown, avoiding a black box in the decision making process. In addition, the computation process can be performed using ordinary PC CPUs because the load on computational resources is small.
◆About IVS LaunchPad
Date: July 31, 2020 (Friday)
Subject of application: Internet, mobile, software, digital contents, consumer electronics (including robots)
Products and services for the Internet and IT in general.
Number of presenting companies: 14
Presentation format: 6-minute demonstration per company (with prior training and mentoring)
Judging process: 1st screening: document review, 2nd and subsequent screenings: presentation, Q&A/mentoring (30-60 minutes)
Evaluation Points: Novelty/uniqueness, marketability/future scalability, presentation skills/composition skills
◆Points of Adacotech’s evaluation (Toshiaki Shimakawa, CEO, Infinity Ventures Summit)
We evaluated Adacotech based on its novelty, uniqueness, and business potential. As many IVS LaunchPad graduates have gone on to IPOs, we look forward to the continued success of Adacotech, the winner of this year’s competition.
* 1HLAC feature extraction method:A general-purpose and fast feature extraction method with excellent recognition accuracy used for image analysis and recognition.In contrast to the Deep Learning technology, which requires complex processing when calculating the shape and size of the inspection target, it can be calculated by simply performing a sum-of-product operation on the values representing the color shade and brightness of each pixel, which can be instantly calculated on a commercial PC.It is also capable of recognizing the same object even if its position has changed, and when there are two objects, the sum of the features of each object becomes the overall feature.In other words, there is no need to find the boundaries of regions that represent the same thing in an image (segmentation free), and the features can be recognized individually even when multiple anomalies occur in an image.It is a feature extraction method with desirable properties for image recognition.