Switzerland

Back to Search

International Partner Search

Innovation & Technology Offer

Greek company developing embedded vision software and IP cores is looking for partners on mobile/portable embedded devices or IoT

Country of Origin: Greece
Reference Number: TOGR20170526001
Publication Date: 29 May 2017

Summary

A Greek based SME develops computer vision software, by using image processing and machine-learning techniques. The SME is currently offering a versatile and easy-to-use software API that allows easy integration of deep learning technology to a variety of modern embedded systems. Ideal partners are those who wish to add computer vision or machine learning capabilities to their devices under financial agreement and/or license agreement.

Description

The Greek company develops embedded vision software and IP cores applied in video stabilization, image signal processing, image fusion, face detection and tracking, license car plate detection and recognition. It addresses the challenge of delivering innovative computer vision solutions while keeping optimal system requirements in terms of power consumption, memory and processing speed.

Their product and technology portfolio include applications in computational photography and visual perception/analytics and address various markets such as mobile devices, action cameras, drones, surveillance, automotive, industrial and robot vision etc.

Moreover, the company offers a versatile and easy-to-use software API that allows easy integration of deep learning technology to a variety of modern embedded systems.

The future brings a wave of embedded devices able to respond to their environment through embedded intelligence. Deep learning is a proven technology able to achieve this intelligence, but requires extremely complex processing to be performed in high speeds and at the same time within low levels of energy consumption.

To achieve those features the company follows a holistic approach:

- Tweak the Models: Use special models and special compression techniques which they are mimicking the human brain and which result in significantly more economical models.

- Tweak the Code: Use heterogeneous programming technologies, featuring the synergistic use of every available computing unit: Multi-cluster CPUs, the GPU and the DSP must all be exploited for carrying out the complex tasks deep learning inference in reasonable time and with limited power budget.

To help developers embed deep-learning technology into their own systems and applications the company developed a software library consisted of high-performance deep-learning models, highly optimized for embedded computing systems and implemented by-using a variety of approaches and techniques, in order to be able to suit to a wide range of applications.

This software includes also the necessary software tools in order to allow the user to train the model of interest by itself and tο use the training results for building the final application.

The company is currently looking for partners that wish to add computer vision or machine learning capabilities to their devices or solutions under financial agreement. Moreover, partners that wish to offer computer vision or machine learning/deep learning capabilities to their customers, are also welcome for cooperation under license agreement.

Advantages and Innovations

The main advantages and innovations of the company: 
- Embedded form
- Low power consumption
- Ease of use
- Personalised solutions

Stage Of Development

Already on the market

Requested partner

The Greek company is looking for cooperation with:
- SMEs or industries active on mobile/portable embedded devices or IoT.
- SMEs or industries that have needs for incorporating computer vision features into their products

Contact Profile Owner