Irida-Labs

Company/Organisation Profiles

Irida Labs is a leading technology provider of software IPs for Embedded Machine Vision using Machine/Deep Learning. The company possesses significant knowledge in analysis, modelling, design and development of high-fidelity reference systems, using state-of-the-art CPU, GPU, FPGA and DSP technologies in the context of heterogeneous computing. Founded in late 2007, Irida Labs is a privately-held company with headquarters in Patras, Greece and worldwide sales support, backed by VC and private investor. Our engineering team is comprised of highly-skilled personnel with MSc/PhD in Machine Vision, guaranteeing the utmost algorithmic effectiveness and efficiency using state-of-the-art methods. Our product and technology portfolio includes embedded vision software for high throughput applications such as object and pattern detection and recognition using Machine/Deep Learning, as well as 3D imaging and reconstruction. Irida Labs has formed collaborations and partnerships with world leaders like Qualcomm, Synopsys, Cadence and Xilinx, and has clients in the USA, China and Europe.

Innovative Products and Services

Irida Labs has developed Va3DPrint, a novel vision-based solution for closing the loop in the Additive Manufacturing (AM) market towards zero-defect AM. This is accomplished by monitoring three different aspects of the AM process/product, namely the 3D geometry of the manufactured parts for quality inspection, the temperature profile for detection of potential cracks, as well as the size, shape and temperature of the melting pool for detecting defects made during the laser deposition process. In addition, Irida Labs has released innovative products based on AI and deep learning for Industry 4.0, Industrial Automation applications and Smart IoT devices such as (1) AEye4, a deep learning-based software tool that integrates AI with embedded vision and heterogeneous computing techniques in order to solve inspections problems with limited set of data and (2) EVLib, a deep learning based software library of highly optimized CNN models that can power AI on edge devices (with functions like Anonymizer – remove sensitive data (GDPR), car/people counting, 2D/3D object detection and tracking etc.). Irida Labs currently has two USPTO patents granted, while a comprehensive set of patent applications have been also filed.

Experience and Expertise

Irida Labs has secured strong R&D funding through National and H2020 projects for its technology in deep learning and 3D vision. Some indicative H2020 projects include:

• “ALOHA –Software framework for runtime-Adaptive and secure deep Learning On Heterogeneous Architectures”, H2020 funded under ICT-05-2017.
• “TEAMPLAY – Time, Energy and security Analysis for Multi/Many-core heterogenous PLAtforms”, H2020 funded under ICT-05-2017.
• “BOREALIS – the 3A energy class Flexible Machine for the new Additive and Subtractive manufacturing on next generation 3D metal parts”, H2020 funded under FoF-02-2014.
• “SYMBIONICA - Reconfigurable Machine for the new Additive and Subtractive Manufacturing of next generation fully personalized bionics and smart prosthetics”, H2020 funded under FoF-10-2015.