Main Reference PaperFitCNN: A Cloud-Assisted and Low-Cost Framework for Updating CNNs on IoT Devices, Future Generation Computer Systems, 2019, [Java]
  • A cloud-assisted CNN framework, named FitCNN is proposed, with incremental learning and low data transmission, to reduce the overhead of updating CNNs deployed on devices. It also proposes a strategy, called Distiller, to selectively upload the data that is worth learning, and develop an extracting strategy, called Juicer, to choose the light amount of weights from the new CNN model generated on the cloud to update the corresponding old ones on devices.

Description
  • A cloud-assisted CNN framework, named FitCNN is proposed, with incremental learning and low data transmission, to reduce the overhead of updating CNNs deployed on devices. It also proposes a strategy, called Distiller, to selectively upload the data that is worth learning, and develop an extracting strategy, called Juicer, to choose the light amount of weights from the new CNN model generated on the cloud to update the corresponding old ones on devices.

  • To reduce the data transmission cost

  • To reduce the uploading transmission cost in FitCNN

  • To improve the accuracy in FitCNN

Aim & Objectives
  • To reduce the data transmission cost

  • To reduce the uploading transmission cost in FitCNN

  • To improve the accuracy in FitCNN

  • A technique is contributed to improve the accuracy further and reduce the data transmission cost.

Contribution
  • A technique is contributed to improve the accuracy further and reduce the data transmission cost.

  • Operating System : Ubuntu 12.04 LTS 64bit

  • Simulator: Cooja, Instant Contiki-3.0 and Vmware Player 12.5.6

  • Language: C

Software Tools & Technologies
  • Operating System : Ubuntu 12.04 LTS 64bit

  • Simulator: Cooja, Instant Contiki-3.0 and Vmware Player 12.5.6

  • Language: C

  • M.E / M.Tech/ MS / Ph.D.- Customized according to the client requirements.

Project Recommended For
  • M.E / M.Tech/ MS / Ph.D.- Customized according to the client requirements.

  • No Readymade Projects-project delivery Depending on the complexity of the project and requirements.

Order To Delivery
  • No Readymade Projects-project delivery Depending on the complexity of the project and requirements.

Professional Ethics: We S-Logix would appreciate the students those who willingly contribute with atleast a line of thinking of their own while preparing the project with us. It is advised that the project given by us be considered only as a model project and be applied with confidence to contribute your own ideas through our expert guidance and enrich your knowledge.

Leave Comment

Your email address will not be published. Required fields are marked *

clear formSubmit