Collaborative Workforce (CW) TeamFall 2011

Advisors:Dr. Randal AblerDr. Edward Coyle

Graduate Student:Kaushiik Baskaran

Under-graduate Students:Scott SimontisChristopher JulianShashank ChamoliDavid Yoon

Outline

• Background and Goals

• Technology Approach

• Engineering Challenges

• Technology Integration

• Current Focus

• Demo

Background & Goals

• Extension of technology in education

• Arbutus Center initiative since 1990’s

• Specific focus on

• Distance Learning

• Telepresence/Teleconferencing

• Video-enabled Collaboration

• A new context is to support technology driven collaborations

• VIP eStadium Team uses CW Team solutions to collaborate

with Purdue & Morehouse students & faculty

Schematic of Van Leer Room 461

Distance Learning Classroom

• Wide use of analog cables

o Variety of analog video formats

o Suffers from noise & attenuation

o Needs amplifiers & converters

o RF interference issues

• Needs an expert to install & maintain the cabling

Cabling

DS

P/A

RM

Technology Approach IP to the end device

Video

Audio

Control

IP over Ethernet

Technology Approach – Video over IP

• System-on-chip (SOC) solution to embedded control & networked media

• On chip media encode & decode capabilities• Incorporates a high-performance TMS320C64x+ DSP

core and an ARM926EJ-S core

• Full HD video over ethernet/IP

• Using existing prototype board at Georgia Tech

Engineering Challenges

• Tradeoffs in media compression technologies• Bandwidth• Delay• Computational complexity• Quality

• H.264 vs. Motion JPEG video

• Device to Device (D2D) delay• Any perceptual delay is challenging

• Room to Room (R2R) delay• Greater delay tolerable

Integration & Convergence

• Single device solution to

• Cabling

• Content switching (Video matrix switch)

• Media digitization & management

• ROI video coding & streaming

• Room control and automation

Current Focus

• Get a feel for how to work with TI’s DaVinci boards

• Create a practical application using the DaVinci boards– Audio Analysis Application

Audio Analysis Application

MATLAB Model

• General example to base C code off of• Can be used to check outputs for errors• Model:

– Input audio file (.wav)– Low Pass Filter (0-11 kHz Equiripple FIR filter)– Decimation (n=2)– 256 point FFT with a Hanning window

Spectrogram

DSP – FFT

ARM sends data to DSP using MGGQ DSPLINK module

DSP does the rigorous MATH and computes the FFT

ARM displays

DSP sends back the FFT Results to ARM using

MSGQ DSPLINK module

ARM - Display

Demo

Questions?