Download presentation

Presentation is loading. Please wait.

Published byKylee Shapley Modified over 3 years ago

1
ISTC-EC @ Cornell Accelerating Belief Propagation in Hardware Skand Hurkat and José Martínez Computer Systems Laboratory Cornell University http://www.csl.cornell.edu/

2
ISTC-EC @ Cornell The Cornell Team Prof. José Martínez (PI), Prof. Rajit Manohar @ Computer Systems Lab Prof. Tsuhan Chen @ Advanced Multimedia Processing Lab MS/Ph.D. students – Yuan Tian, MS ’13 – Skand Hurkat – Xiaodong Wang

3
ISTC-EC @ Cornell The Cornell Graph

4
ISTC-EC @ Cornell The Cornell Project Provide hardware accelerators for belief propagation algorithms on embedded SoCs (retail/car/home/mobile) – High speed – Very low power – Self-optimizing – Highly programmable BP Accelerator within SoC Graph Inference Algorithm Result

5
ISTC-EC @ Cornell What is belief propagation? Belief propagation is a message passing algorithm for performing inference on graphical models, such as Bayesian networks or Markov Random Fields

6
ISTC-EC @ Cornell What is belief propagation? Labelling problem Energy as a measure of convergence Minimize energy (MAP label estimation) Exact results for trees – Converges in exactly two iterations Approximate results for graphs with loops – Yields “good” results in practice Minimum over large neighbourhoods Close to optimal solution

7
ISTC-EC @ Cornell Not all “that” alien to embedded Remember the Viterbi algorithm? Used extensively in digital communications

8
ISTC-EC @ Cornell What does this mean? Every mobile device uses Viterbi decoders – Error correction codes (eg: turbo codes) – Mitigating inter-symbol interference (ISI) Increasing number of mobile applications involve belief propagation – More general belief propagation accelerators can greatly improve user experience with mobile devices

9
ISTC-EC @ Cornell Target markets Retail/Car/Home/Mobile Image processing – De-noising – Segmentation – Object detection – Gesture recognition Handwriting recognition – Improved recognition through context identification Speech recognition – Hidden Markov models are key to speech recognition Servers Data mining tasks – Part-of-speech tagging – Information retrieval – “Knowledge graph” like applications Machine learning based tasks – Constructive machine learning – Recommendation systems Scientific computing – Protein structure inference

10
ISTC-EC @ Cornell Hardware accelerator for BP BP Accelerator within SoC Graph Inference Algorithm Result

11
ISTC-EC @ Cornell Work done so far

12
ISTC-EC @ Cornell Work done so far

13
ISTC-EC @ Cornell Work done so far

14
ISTC-EC @ Cornell Hierarchical belief propagation

15
ISTC-EC @ Cornell Results – Stereo Matching

16
ISTC-EC @ Cornell Work done so far

17
ISTC-EC @ Cornell Work done so far

18
ISTC-EC @ Cornell GraphGen synthesis of BP-M

19
ISTC-EC @ Cornell Cornell Publications (2013 only) 3x Comp. Vision & Pattern Recognition (CVPR) 3x Asynchronous VLSI (ASYNC) 2x Intl. Symp. Computer Architecture (ISCA) 1x Intl. Conf. Image Processing (ICIP) 1x ASPLOS (w/ GraphGen folks, under review)

20
ISTC-EC @ Cornell Year 3 Plans GraphGen extensions for BP applications – Multiple inference techniques Extraction of “BP ISA” – Ops on arbitrary graphs – Efficient representation Amplification work on UAV ensembles – Self-optimizing, collaborative SoCs One-day “graph” workshop with GraphGen+UIUC

21
ISTC-EC @ Cornell Accelerating Belief Propagation in Hardware Skand Hurkat and José Martínez Computer Systems Laboratory Cornell University http://www.csl.cornell.edu/

22
ISTC-EC @ Cornell Where can it be used? Image processing – Stereo matching – Image segmentation – Identifying objects in context Protein structure inference Almost any algorithm that uses Markov models – Speech recognition using HMM – Handwriting recognition

23
ISTC-EC @ Cornell Graphgen generator for (BP) apps BP Generator Applications Stereo Segmentation etc Algorithms BP-M Hierarchical etc GraphGen Spec Accelerators

24
ISTC-EC @ Cornell The math

25
ISTC-EC @ Cornell The math is not so alien or Remember the Viterbi Algorithm?

26
ISTC-EC @ Cornell The math is not so alien Let or

27
ISTC-EC @ Cornell The math is not so alien The Viterbi algorithm is merely a simpler version of belief propagation!

28
ISTC-EC @ Cornell What does this mean? Servers can also benefit from BP accelerators – Data mining tasks Part-of-speech tagging Information retrieval “Knowledge graph” like applications – Machine learning based tasks Constructive machine learning Recommendation systems – Scientific computing Protein structure inference

Similar presentations

OK

Discriminative Training and Machine Learning Approaches Machine Learning Lab, Dept. of CSIE, NCKU Chih-Pin Liao.

Discriminative Training and Machine Learning Approaches Machine Learning Lab, Dept. of CSIE, NCKU Chih-Pin Liao.

© 2018 SlidePlayer.com Inc.

All rights reserved.

Ads by Google

Ppt on waves tides and ocean currents indian Ppt on supply chain management of nokia cell Ppt on hotel industry in india 2013 Run ppt on html editor Ppt on ms word 2007 tutorial Ppt on depth first search algorithm Ppt on lathe machine parts Ppt on boilers operations management Ppt on bluetooth communication devices Ppt on network security issues