Presentation is loading. Please wait.

Presentation is loading. Please wait.

Pradeep K. Dubey, Applications Pradeep K Dubey for Bob Liang Applications Research Lab Microprocessor Technology Labs Corporate.

Similar presentations


Presentation on theme: "Pradeep K. Dubey, Applications Pradeep K Dubey for Bob Liang Applications Research Lab Microprocessor Technology Labs Corporate."— Presentation transcript:

1 Pradeep K. Dubey, pradeep.dubey@intel.com Applications Pradeep K Dubey for Bob Liang Applications Research Lab Microprocessor Technology Labs Corporate Technology Group December 8, 2005

2 Pradeep K. Dubey, pradeep.dubey@intel.com 2 Application Research Lab Carole DulongPradeep Dubey Tim Mattson Gary Bradski Horst Haussecker Jim Hurley Bob Liang

3 Pradeep K. Dubey, pradeep.dubey@intel.com 3 What is a killer app?  “A reasonable man adapts himself to his environment. An unreasonable man persists in attempting to adapt his environment to suit himself …  … Therefore, all progress depends on the unreasonable man.” -- George Bernard Shaw  Replace “man” with “application”, and you get one definition of a killer app, namely that unreasonable application which succeeds in leaving its mark on the surrounding architecture. All architectural progress depends on such unreasonable apps!

4 Pradeep K. Dubey, pradeep.dubey@intel.com 4 Mining Indexing Recognition Synthesis Streaming Graphics Large dataset mining Semantic Web/Grid Mining Streaming Data Mining Distributed Data Mining Content-based Retrieval Collaborative Filters Multidimensional Indexing Dimensionality Reduction Dynamic Ontologies Efficient access to large, unstructured, sparse datasets Stream Processing Multimodal event/object Recognition Statistical Computing Machine Learning Clustering / Classification Model-based: Bayesian network/Markov Model Neural network / Probability networks LP/IP/QP/Stochastic Optimization Photo-real Synthesis Real-world animation Ray tracing Global Illumination Behavioral Synthesis Physical simulation Kinematics Emotion synthesis Audio synthesis Video/Image synthesis Document synthesis Data-data everywhere, not a bit of sense!

5 Pradeep K. Dubey, pradeep.dubey@intel.com 5 Scene complexity: moderate Local processing dominated Media Evolution Graphics Evolution Mining Evolution Scene complexity: large Global processing dominated Scene complexity: real-world Physical simulation dominated Modality-specific streaming Modality-aware transformation Multimodal recognition Dataset: static/structured Response: offline Dataset: dynamic, multimodal Response: real-time Dataset: massive+streaming Response: interactive Upcoming Transition Next Transition Evolving towards model-based computing Workload convergence: multimodal recognition and synthesis over complex datasets

6 Pradeep K. Dubey, pradeep.dubey@intel.com 6 What is a tumor?Is there a tumor here? What if the tumor progresses? It is all about dealing efficiently with complex multimodal datasets RecognitionMiningSynthesis Images courtesy: http://splweb.bwh.harvard.edu:8000/pages/images_movies.html

7 Pradeep K. Dubey, pradeep.dubey@intel.com 7 Visual Input Streams Find an existing model instance What is …?What if …? Is it …? RecognitionMiningSynthesis Create a new model instance What Killer app – grep, ctrl-c, ctrl-v? Model Most RMS apps are about enabling interactive (real-time) RMS Loop or iRMS Synthesized Visuals Learning & Modeling Graphics Rendering + Physical Simulation Computer Vision Reality Augmentation

8 Pradeep K. Dubey, pradeep.dubey@intel.com 8 iRMS Loop Illustration Facial Muscle Activations: Compact motion representation, well suited for modeling and synthesis Video InputFeature Tracking Analytically Correct, Muscle- Activated Human Head Model Physics-Based Deformable Tissue (Finite Element Method) User Interaction: Modified Muscle Activations Video Output User Interaction: Modified Physical Model Source: E. Sifakis, I. Neverov and R. Fedkiw, “Automatic Determination of Facial Muscle Activations from Sparse Motion Capture Marker Data”, ACM SIGGRAPH, 2005 (to appear)

9 Pradeep K. Dubey, pradeep.dubey@intel.com 9 Virtual Reality, Games, Simulations …

10 Pradeep K. Dubey, pradeep.dubey@intel.com 10 Computer Vision – OpenCV 1M downloads! Video Camera Input Desert Road identified Parallel Body Tracker Adding vision input as Natural Interface to Physics and Rendering Visual Programming – rendering, physics, vision input

11 Pradeep K. Dubey, pradeep.dubey@intel.com 11 Visual Computing Closing the loop between computer vision, physical simulation, and photo-realistic rendering, and building an interactive system. Tracking Physics Simulation Multi-camera input Rendering Applications: Games Virtual Reality Surgery and health care Virtual dressing room Movies and special effects... Applications: Games Virtual Reality Surgery and health care Virtual dressing room Movies and special effects...

12 Pradeep K. Dubey, pradeep.dubey@intel.com 12 Digital Libraries in the 90’s  Data Base extenders for media data management  Server based  CBIR  IBM QBIC  Virage, etc.  Good for Ad professional  Similarity for fade, wipe, etc  Consumers want  “just find it”  Natural user interface

13 Pradeep K. Dubey, pradeep.dubey@intel.com 13 Demos

14 Pradeep K. Dubey, pradeep.dubey@intel.com 14  New Killer App?:  There isn’t one -- Same old one:  There isn’t one -- Same old one: grep, ctrl-c, ctrl-v  It’s a parallel world!  Shall we look on the other side of the serial death valley?  It’s an analog and non-linear world!  Computers have digitized and linearized, but … - Real-world problems are still largely non-linear and analog  Almost infinite appetite for computational power, if … - You reach a certain threshold needed for simulated interactions in real- time. Summary

15 Pradeep K. Dubey, pradeep.dubey@intel.com 15 Thank You!

16 Pradeep K. Dubey, pradeep.dubey@intel.com 16 Back-up

17 Pradeep K. Dubey, pradeep.dubey@intel.com 17 Tomorrow What is …?What if …? Is it …? RecognitionMiningSynthesis Create a model instance RMS: Recognition Mining Synthesis Model-based multimodal recognition Find a model instance Model Real-time analytics on dynamic, unstructured, multimodal datasets Photo-realism and physics-based animation Today Model-lessReal-time streaming and transactions on static – structured datasets Very limited realism

18 Pradeep K. Dubey, pradeep.dubey@intel.com 18 Visual Input Streams Find an existing model instance What is …?What if …? Is it …? RecognitionMiningSynthesis Create a new model instance Emerging Workload Focus: iRMS Model Most RMS apps are about enabling interactive (real-time) RMS Loop or iRMS Synthesized Visuals Learning & Modeling Graphics Rendering + Physical Simulation Computer Vision Reality Augmentation

19 Pradeep K. Dubey, pradeep.dubey@intel.com 19 “Cognitive SQL”? SQL Today: Table USER: “Give me all employees who make between $80K and $87K” SQL API Dave, 37, 7, $82,000, …... USER: “Find the variable that most describes compensation” Total Compensation Tenure with Company IPO Cognitive SQL : Spectral Clustering Probabilistic Models Inference Clustering, Feature Selection Histograms Discriminative Classifiers/Trees ML API SQL API

20 Pradeep K. Dubey, pradeep.dubey@intel.com 20 Summary  Design parallel algorithms with parallel computing mindset from the beginning, not parellelizing serial algorithms. Even “inherently” parallel applications such as Ray tracing and computer vision requires work  Potential killer app - To satisfy consumer’s requirement of “Just Find it” with natural user interface  Examples of (iRMS) Interactive Recognition-Mining- Synthesis – the essence is the timely delivery of the knowledge  Machine learning techniques will play an important role in help us extract useful knowledge from the massive amount of digital dataset  Explore the parallel programming patterns for each domain. before we have a book “Parallel computing for dummies”.

21 Pradeep K. Dubey, pradeep.dubey@intel.com 21 the algorithms can be re-formulated as a sum over the data points and the sum can be broken up over one to many threads Algorithm Have summation form? 1. Linear regression Yes 2. Locally weighted linear regression Yes 3. Logistic regression Yes 4. Gaussian discriminant analysis Yes 5. Naïve Bayes Yes 6. SVM (without kernel) Yes 7. K-means clustering Yes 8. EM for mixture distributions Yes 9. Neural networks Yes 10. PCA (Principal components analysis) Yes 11. ICA (Independent components analysis) Yes 12. Policy search (PEGASUS) Yes 13.BoostingUnknown 14. SVM (with kernel) Unknown 15. Gaussian process regression Unknown Machine Learning on Multi-Core

22 Pradeep K. Dubey, pradeep.dubey@intel.com 22 More powerful computer to help us discover new knowledge? Computer has been used to help Researchers discover new knowledge “Pure Mathematics” - 4 color problem We have computational geometry, computational chemistry, etc. Will we have computational history?

23 Pradeep K. Dubey, pradeep.dubey@intel.com 23 Applications “New” Innovative Applications? “There is nothing new under the sun” This talk: multi-core processing power and “new” techniques to make some “old” applications work Innovative Applications -> Afternoon Panel


Download ppt "Pradeep K. Dubey, Applications Pradeep K Dubey for Bob Liang Applications Research Lab Microprocessor Technology Labs Corporate."

Similar presentations


Ads by Google