Deep Learning

Our Know-How for Your Ideas

Correlance offers state of the art science-based solutions for small, medium and large companies that need advanced Research & Development services, primarily in the following markets:

  • Audio-Video Broadcast Analysis: real time systems for semantic analysis of audio-video broadcast. Logo Detection for advertising measurement, certification of Commercial Breaks.
  • Deep Learning for Data Modeling: Deep Learning modeling oriented to predictive analysis
  • Industrial Automation: quality control, statistical classification and predictive maintenance systems.

Mathematical & Scientific Analysis

Analysis of problems and research of state of the art science-based solutions, through the synergy of the following subjects:

  • Digital Image Processing
  • Digital Audio Processing
  • Inverse Problems (linear & non linear)
  • Optimization Problems
  • Statistical Learning
  • Pattern Recognition

Scientific coding

Rapid prototyping using computer science, mathematical and computer vision tools:

  • Efficient math computing (Lapack, Blas)
  • Intensive math computing (GPCPU: OpenCL, OpenGL, CUDA)
  • Audio-video codec (DirectShow, LibAV)
  • Computer Vision (OpenCV)
  • Scientific prototyping (Matlab)

Efficient coding

Developing effective code implementation, taking care of computational efficiency and memory footprint:

  • Object oriented programming (C++, C#)
  • Parallel programming (OpenMP, TBB)
  • Integer and fixed point coding (ARM, DSP)
  • Code optimization (SSE2, SSE3, SSE4, assembler)
  • Distributed programming (.Net remoting)

Project Management

Correlance team has years of experience in advanced and complex Research and Development projects. The technical team follows an agile approach for the design process, with milestones, that consist oh these following steps:

  • problem analysis
  • prototype requirements
  • prototype development iterations
    • prototype design
    • prototype development
    • prototype test
  • product requirements
  • product development iterations
    • product design
    • product development
    • product test
  • acceptance test
  • corrective maintenance and new features implementation

Particular attention is paid to scientific documentation, useful to any practice of industrial property rights, and to product documentation.