Interreg - DeepVis: Visualization of deep learning neural network to support the utilisation of artificial intelligence by SMEs

REQUEST from United Kingdom, reference: RDUK20180302003, valid from 02-03-2018 untill 23-03-2018

Research & Development
  • Start date:
    2 maart 2018
  • End date:
    23 maart 2018
  • Summary:
    An East of England based University are developing a proposal for the forthcoming Interreg call. The proposal, DeepVis, is designed to enhance the innovation performance of SMEs in North West Europe (NWE) region by supporting them in the utilisation of deep learning tools (DL) and resources to leverage Artificial Intelligence (AI) in their core business through visualisation. A number of partners are required to complete the project consortium.
  • Description:
    A University in the East of England has developed visualisation tools that allow less technical SMEs implement modern AI and DL tools for higher productivity and competitiveness. They will submit a proposal called DeepVis to the Interreg North West Europe programme.
    AI products like Alexa, Siri, Cortana, and Google create huge market value. Some current major uses of AI include image recognition, object identification, detection, and classification, automated geophysical feature detection, and so on. According to report, by 2020, big & open data can improve the European GDP by 1.9%, with positive impact on employment. Studies also suggest that companies that adopt big data analytics can increase productivity by 5%- 10% more than companies that do not.

    Investigating the potential impact of Big Data in SMEs is the key for the growth in both micro (firm-based) and macro (economy) levels. While AI may give small businesses a huge boost, there are significant barriers for the SMEs to leverage the latest progress in AI, despite the promises about the potential of AI. A recent survey shows that 61% of small business owners believe their businesses are not ready for AI; 58% of businesses have considered AI technologies, but only 12% are actually using these systems, leaving a huge opportunity for early adopters to pull ahead of their competitors.

    The most important development in deep learning (DL) in recent years has been the marked increase in performance in many applications if their capacities are harnessed properly. However, DL demands very specific knowledge and skills to ensure that its potential is fully realized – it often involves ever-large networks, which is very hard to comprehend how they operate, and why they are so powerful, making it difficult to test, adjust and improve the outcomes.

    Thus, methods for visualizing the decision-making process and inner workings of deep neural networks can be of great value for their qualitative assessment. Understanding them better will enable the developers to find new ways to guide training into the right direction and improve existing successful networks by detecting their weaknesses, as well as accelerating their adoption in applications.

    The overall benefit of the DeepVis toolkit for the SMEs includes:
    • Easy utilisation of the latest AI/DL outcomes/resources to achieve significant performance gains in their products through a friendly UI without requiring in-depth knowledge
    • Improved understanding and interpretation of the outcomes from DL in order to help the developers maximise the performance
    • Good interoperability between the toolkit and major DL programming languages to allow the developers to easily integrate the outcome from the toolkit into their current work.

    A number of partners are required to complete the proposal consortium - shown below:

    SME associations from all North West Europe (NRW) region countries
    • SMEs with data analytics as core business
    • Policy makers for economic development and SMEs from all NWE region countries
    • DL/AI developers from all NWE region countries to co develop the tool.

    A number of core SMEs and SME associations in terms of carrying out detailed requirement analysis, co-design and early test of the solution on their product. The policy makers will also provide very useful insights into the problem.

    The tool will need to be tested on a few pilot sites across NWE, involving 10 – 20 SMEs as the direct beneficiaries of the technology. The core SMEs and SME associations onboard is expected to help identify the pilot sites.

    The core SME associations and policy-making partners in the consortium to contribute to the dissemination to attract a large number of SMEs to adopt the toolkits in their product development.

    EOI deadline - 23rd March 2018
    Call Deadline - 26th April 2018
Partner Sought
  • Type of Partnership Considered:
    Research cooperation agreement
  • Technical Specification or Expertise Sought:

    The consortium expects the outcome of the project to bring increase of the sales/profits/income in many SMEs in near future (i.e 5 to 10 years after the completion of the project), with increased SME innovation levels, and enhanced capacity of NWE regions and territory for joint research & innovation.

    To achieve this, the consortium will need significant input from different key stakeholders. Hence they invite participations especially from:
    • SME associations from all NWE region countries
    • SMEs with data analytics as core business
    • Policy makers for economic development and SMEs from all NWE region countries
    • DL/AI developers from all NWE region countries to co develop the tool.

    Note the North-West Europe include the following countries: United Kingdom, Belgium,
    Luxembourg, Switzerland and and parts of France, Germany and the Netherlands.

    The group needs a number of core SMEs and SME associations in terms of carrying out detailed requirement analysis, co-design and early test of the solution on their product. The policy makers will also provide very useful insights into the problem.

    In addition, the tool will need to be tested on a few pilot sites across NWE, involving 10 – 20 SMEs as the direct beneficiaries of the technology. The core SMEs and SME associations onboard is expected to help identify the pilot sites.

    Finally, strong commitment to the promotion of the project outcomes will be another key to success. Similarly, the group expects the core SME associations and policy-making partners in the consortium to contribute to the dissemination to attract a large number of SMEs to adopt the toolkits in their product development.

    SME 11-50,R&D Institution,SME <10,SME 51-250

Organisation
  • Development stage:
    Proposal under development
  • IPR Status:
  • Market keywords:
    Data processing, analysis and input services
    Big data management
    Databases and on-line information services
    Business and office software
    Artificial intelligence related software
  • Technology keywords:
    Artificial Intelligence (AI)
    Computer Software
    Databases, Database Management, Data Mining
    Electronic Commerce, Electronic Payment & Signature
  • NACE keywords:
  • Advantages and Innovations:
Client
  • Type and Size of Organisation:
    University
  • Already Engaged in Trans-National Cooperation
    Nee
  • Year Established:
    0
  • Turnover:
  • Country of origin:
    United Kingdom
Program
  • Coordinator Required
    Nee
  • Deadline:
    26 april 2018
  • Submission and evaluation scheme: