IEEE ESPA 2012

susie wee @susiewee

Augmented reality: Enabled by scene estimation, 3D object reco, projective display, context. Dr. Mike Polley, TI Fellow #ESPA12

Daniele Giacobello @dgiacobello

#dsp RT @susiewee: Applications: Gesture-based gaming, Gesturing in cars, Interactive design. Dr. Mike Polley, TI Fellow #ESPA12 #in

susie wee @susiewee

Applications: Gesture-based gaming, Gesturing in cars, Interactive design. Dr. Mike Polley, TI Fellow #ESPA12

susie wee @susiewee

Vision library elements: Background modeling & subraction, Feature extraction, Low-level pixel processing, Tracking and recognition. #ESPA12

susie wee @susiewee

Dr. Mike Polley, TI Fellow: Developer can use Computer Vision libraries to get HW accelerations. HW gives 10x perf improvement #ESPA12

susie wee @susiewee

Dr. Mike Polley, TI Fellow, discussing the evolution of algorithms from app level, down to HW, down to chips. Need to provide APIs #ESPA12

Daniele Giacobello @dgiacobello

lot of gesture recognition at #ces and already a big component of the #espa12 talks. Looking forward to see it on a smartphone/tablet

susie wee @susiewee

Dr. Mike Polley, TI Fellow: Need to provide an app programming paradigm that allows access to the HW acceleration. #ESPA12

susie wee @susiewee

Dr. Mike Polley, TI Fellow: 80/20/80 rule of thumb. 80% of vision app runs on fully prog processor. 20% of code on dedicated HW. #ESPA12

Minoru Mick Etoh @mickbean

RT @susiewee: Next: "Signal Processing and Big Data in Mobile Communication" by DoCoMo Innovation CEO @mickbean at #ESPA12 http://t.co/vnfHctyL

Minoru Mick Etoh @mickbean

RT @susiewee: Next up at #ESPA12 is @mickbean , Dr. Minoru Etoh, VP of NTT DoCoMo and CEO of DoCoMo Innovations USA. Looking forward to his insights!

susie wee @susiewee

Dr. Mike Polley, TI fellow, discussing Sensor Fusion algorithms. #ESPA12

susie wee @susiewee

Dr. Mike Polley showing a picture of cell phones with linear arrays of cameras. He's showing cell phones with lots of sensors! :-) #ESPA12

susie wee @susiewee

Dr. Mike Polley discussing the multiplication of sensors: Increased # per devices, increased quality, reduced power. #ESPA12

susie wee @susiewee

Multimedia processing increase will flatten out, but Computer Vision processing will increase by 2020. Dr. Mike Polley, TI #ESPA12

susie wee @susiewee

Dr. Mike Polley of TI: 2000: ARM7 QCIF 30, MP4. 2010: 1080P30 H.264HP. 2020: 2Kx4K #ESPA12

susie wee @susiewee

Dr. Mike Polley TI: Mobile App processor trends 4) New apps on programmable cores move to dedicated cores #ESPA12

susie wee @susiewee

Dr. Mike Polley TI: Mobile App processor trends 3) Domain-specific cores (multimedia) #ESPA12

susie wee @susiewee

Dr. Mike Polley TI: Mobile App processor trends 1) Thermal limit is a hard barrier 2) Programmable cores - CPU, DSP, GPU, etc. #ESPA12

susie wee @susiewee

Dr. Mike Polley on Harnessing Embedded Signal Processing to Enhance User Experiences on Future Mobile Devices. http://t.co/k6gAchu3 #ESPA12

Daniele Giacobello @dgiacobello

Shift: aggregated data as signals --> challenge: integrate platforms for "Big Data Understanding". Dr. Minoru Etoh, NTT Docomo. #espa12 #in

susie wee @susiewee

Great talk by Dr. @mickbean at #ESPA12 . Very interesting insights in a set of apps I don't think about everyday.

susie wee @susiewee

Opportunity: Aggregated data as signals. Challenge: Integrated platforms for big data understanding. Dr. @mickbean #ESPA12

susie wee @susiewee

Big data intelligence (2011) = data+machine learning (2005) + Web-scale data aggregation (2006) + massively parallel processing #ESPA12

susie wee @susiewee

After earthquake data usage declined significantly. Then declined more as cell phone batteries ran out. Dr. @mickbean #ESPA12