Why Retail Business Needs Image Recognition Technology Today.
It’s the game-changing technology that big businesses simply can’t operate without. In today’s cut-throat global retail marketplace, AI & image recognition technology is paramount.
This progress though brings its own challenges, such as supply of sufficient talent, resources, and data. The most pressing problem, is the availability of well-labeled data to support the supervised learning process. Even when unstructured data is abundant, converting it to a labeled dataset is an extremely laborious and costly process. There is a need to address the challenges of practical AI adoption wholesale — on a platform level. Until this is done, most businesses will be struggling to adopt AI. Due to the disparity in resources and focus among companies, we will continue to see a widening gap between early adopters and legacy operators, putting the latter at an extreme strategic disadvantage.
The latest Euro money country risk survey conducted among 72 of the largest retailers in Europe shows the monthly need for image recognition in the global retail market comprises of at least 1.9 billion photos, with a projected growth of up to 3.5 billion over the next two years.
The potential demand for image recognition alone in the retail industry is enormous. By 2020, it’s expected that 85 per cent of customer interactions in retail will be managed by artificial intelligence.
Until recently, the problem for retailers has been a lack of technology for the real-time visual recognition of goods, but that’s all about to change. Joint trade and industry body, Efficient Consumer Response, has partnered with Neuromation.io, creating a revolutionary solution for businesses. Neuromation.io’s solution for retail was presented at ECR’s September forum in Milano, showcasing several products that use the computing power of mining equipment to generate synthetic data to grow traditional retail. This technology is set to be integrated into ECR’s service, giving merchandisers a clear understanding of the presence, absence and layout of goods on the shelf in real-time.
A key advisor and mentor to the Neuromation.io team on object recognition is one of the creators of Google image and Tensor Flow, Andrew Rabinovich. As a world-leading scientist for deep learning and image recognizing research, Andrew has been studying machine learning with an emphasis on computer vision for over 15 years. He’s also the founder of a biotechnology startup and the author of numerous patents and peer-reviewed publications.
The technology developed by the team at OSA Hybrid platform is already well-established in Eastern Europe and Russia, where it was implemented by ECR into large retail chains & companies such as Metro Cash & Carry, Auchan, Magnit, X5 Retail Group, Dixy Group, Verny, Monet, Rainbow Smile, PepsiCo, Danone, Mars, Coca-Cola, Unilever, L’Oréal, SunInBev, JTI and Efes. By processing this data in real-time, the algorithms embedded in the platform allow for instant and high-precision detection of not only the absence of goods on the shelf, but to the cause, sending a signal about the problem and recommendations for its resolution to the desired part of the supply chain.
In Russia, the ECR has gone a step further, using Neuromatiom.io’s technology to create media content. This content is used by participating companies of the ECR Russia ECO System in internal processes, as well as for training neural networks. By the time the product hits the shelf, the OSA HP system in conjunction with image recognition, is already able to recognize the goods and collect all relevant information on the availability of the goods on the shelf… along with prices, location relative to planograms and control of promotion calculations.
Russia’s ECR Executive Director, Maximilian Musselius says “The introduction of image recognition technology will increase coverage to 100 per cent of sales points, automate the recognition process, reduce processing time and errors, and extend the technology to the level of the industry standard.
Neuromation.io has also declared the possibility of implementing object recognition on end-user devices (merchandisers) without the need to connect to a server, which will allow rapid scaling of ECR services. Neuromation.io set the target for accuracy of recognition at 95 per cent.” Musselius went on to say that “Retailers will now have an additional opportunity to increase retail turnover by 3.5 per cent per annum, by integrating Image Recognition Neuromation.io technology into the ECR OSA hybrid platform. This will give an additional effect to the already existing 5.4 per cent sales growth when using the OSA HP service.”
The developer behind the platform, Valentin Ovechkin, says these results are significant because “We proved that there is a real application of big data and artificial intelligence technologies in offline retail. The service works in real-time and allows the supplier and retailer to optimize costs and boost sales simultaneously.”
The ECR will implement this technology in the platform already used by its participants, along with companies and members of the ECR community.
The service is free for the retailer, despite the fact that the retailer gains sales growth.
With full-scale implementation of the service being rolled out until the middle of 2018, Neuromation.io is inviting potential investors to get onboard now. Exponential growth is projected to continue in the store inventory logistics robot market, providing an opportunity for ripe investment over the coming 15 years.
In order to transact on Neuromation’s platform, a client will need to buy tokens. To simplify the purchase mechanics, Neuromation will provide a client portal that makes token purchase a one-click process.
Neurotokens for the ICO go on sale on October 15.
For more information visit the Neuromation.io website
