Convolution Neural Networks

Convolutional Neural Networks (CNNs) are a type of Neural Networks that have proven very effective in areas related to classification and categorization -  for example, image & speech recognition. CNNs are applicable for multiple verticals ranging from Autonomous Driving to Retail Analytics to Security and Surveillance systems.

While CNNs hold a lot of promise, very few companies actually know how to use them and they bring with them a lof of hype and undelivered promises. 

HSC has been working on the application of CNN as a subset of its Machine Learning skillsets specifically in the following areas:

  • Application of CNN modeling for autonomous driving
  • Retraining CNN layers for alternate product classification (example, retraining a model for identifying faces to identify product packaging instead)
  • How to leverage GPU and GPU clusters for faster classification (including fully contained device classification without a cloud backend)

As of today, HSC works with Digital Innovation labs of Retail companies, Security and Surveillance and IoT OEMs in the area of CNN engineering where we help create and optimize different CNN models.

CNN Frameworks HSC typically works on:

  • Tensorflow
  • Caffe 
  • IBM Watson

Contact Us

 

Disclaimer: Hughes Systique shall not be liable for any loss or damage sustained by reason of any disclosure (inadvertent or otherwise) of any information concerning the user's account nor for any error, omission or inaccuracy with respect to any information so disclosed and used whether in pursuance of a legal process or otherwise.

Any other personal information (including sensitive personal information) shared by you which is not asked by Hughes Systique during registration, either mandatorily or optionally; accounts to wilful and intentional furnishing; and Hughes Systique will not be liable for breach of such information.

Related pages and insights