
Deep Block will showcase its AI platform at the World IT Show 2023 which will take place from April 19th - 21st at the COEX convention center, Seoul, South Korea, booth BB100. Deep Block is an AI platform that helps GIS analysts and data scientists easily analyze large amounts of remote sensing imagery using state-of-the-art machine learning algorithms.
Omnis Labs Co., Ltd., an AI venture company founded in 2019, provides AI development services. Omnis Labs created its own no-code AI platform - Deep Block - with the aim of creating a world where anyone can adopt artificial intelligence technology. Specializing in geospatial data analysis such as satellites and aerial images, Deep Block provides a drag-and-drop-based UI that allows users to train and use high-performance machine learning-based image analysis models in just a few hours, thanks to its library of pre-trained algorithms.
Despite the fact that most organizations and specialists are currently analyzing massive amounts of geospatial data by relying on human labor, deep learning technologies have shown their capabilities to automate up to 70% of menial analysis tasks. However, due to the large amount and resolution of remote sensing imagery, it is often difficult for these organizations properly leverage AI technology. Additionally, there is a high demand for AI experts, making it challenging for most organizations to assemble their own research teams. Furthermore, most geospatial industry workers and research personnel lack expertise in AI, as they typically have backgrounds in surveying, urban engineering, and physics.
To address these challenges, Deep Block provides an industrial-grade development pipeline for geospatial models and a library of pre-trained algorithms. For instance, Deep Block offers an object recognition model that can detect ground objects in satellite, aerial, and drone images and analyze their shape, size, location, or changes over time. This solution is beneficial for homeland security and defense agencies for a growing variety of use cases, such as ground, marine, and strategic location monitoring. Deep Block is also useful for other use cases such as urban planning, environmental monitoring, and disaster management.
Recently, Omnis Labs Co., Ltd. announced the development of ultra-high-capacity satellite image processing technology:
- 15GB image files support (png, jtif, jpg, tiff, bmp, jp2)
- 120,000 x 120,000 pixel max. image resolution support
- High data processing rate: 15GB/sec for object recognition
- High accuracy object detection algorithm: 0.9 mAP on average, 0.9 Recall
Omnis Labs Co., Ltd. plans to continue to provide various AI-based solutions to researchers and workers in the geospatial industry, but also in the public and defense fields. The company also looks forward to partnering with various companies in the industry, with the goal of further developing the world's best technology that can unlock remote sensing imagery analysis at scale and democratize artificial intelligence technology.