Adrian Barbu - Homepage

   Adrian Barbu, Assistant Professor, Department of Statistics, Florida State University.
  
   Research interests:
       - Learning Based Modeling and Computing
       - Computer Vision
       - Hierarchical Computing
       - Medical Imaging

Teaching:

• Fall 2007, 2008,2009: Applied Linear Regression, STA 4203/5207
• Spring 2008,2009: Applied Machine Learning
• Spring 2010: Medical Image Analysis, STA 4930-1/STA5934-1

Journal Publications

1. F. Bunea, A. Tsybakov, M. Wegkamp and A.Barbu. SPADES and mixture models. To appear in Annals of Statistics.
2. F. Bunea and A.Barbu. Dimension reduction and variable selection in case control studies via regularized likelihood optimization.
   Electronic Journal of Statistics, 3, 2009. (pdf)
3. A. Barbu. Training an Active Random Field for Real-Time Image Denoising. IEEE Trans. Image Processing, 18, November 2009. (pdf)
4. Y. Zheng, A. Barbu, B. Georgescu, M. Scheuering and D. Comaniciu. Four-Chamber Heart Modeling and Automatic Segmentation
   for 3D Cardiac CT Volumes Using Marginal Space Learning and Steerable Features. IEEE Trans Medical Imaging, November 2008. (pdf)
5. A. Barbu, S.C. Zhu. Generalizing Swendsen-Wang for Image Analysis. J. Comp. Graph. Stat. 16, No 4, 2007 (pdf)
6. A. Barbu, S.C. Zhu. Generalizing Swendsen-Wang to sampling arbitrary posterior probabilities, PAMI, 27, August 2005 (pdf)
7. C.V. Ciobanu, A. Barbu, R.M. Briggs. Interactions of carbon atoms and dimer vacancies on the Si(001) surface.
   Journal of Engineering Materials and Technology -ASME 127, 462 (2005) (pdf) 
8. A. Barbu. On the range of non-vanishing p-torsion cohomology for GLn(Fp), Journal of Algebra, 278, pp 456-472, August 2004 (pdf, link)
9. A. Barbu. On a conjecture of Ash, Journal of Algebra, 251, pp 178-184, May 2002 (pdf, link)
10. A. Barbu. The ring generated by the elements of degree 2 in H*(Un(Fp),Z ), Journal of Algebra, 237, pp 247-261, March 2001 (pdf, link)
11. Cluster sampling and its application to segmentation, stereo and motion (PhD thesis, UCLA 2005) (pdf)
12. On the cohomology of GLn(Fp) with Fp coefficients (PhD thesis, OSU 2000) (pdf)

Conference Publications

1. A. Barbu. Learning Real-Time MRF Inference for Image Denoising. CVPR 2009 (pdf)
2. A. Barbu, R. Ionasec. Boosting Cross-Modality Image Registration. URBAN 2009 (pdf)
3. S. Seifert, A. Barbu, S. Zhou, D. Liu, J. Feulner, M. Huber, M. Suehling, A. Cavallaro, D. Comaniciu. 
   Hierarchical parsing and semantic navigation of full body CT data. SPIE Medical Imaging, 2009 (pdf)
4. L. Lu, A. Barbu, J. Liang, L. Bogoni, M. Salganicoff and D. Comaniciu. Simultaneous Detection and Registration
   for Ileo-Cecal Valve Detection in 3D CT Colonography. ECCV 2008 (pdf)
5. L. Lu, A. Barbu, M. Wolf, J. Liang, M. Salganicoff, D. Comaniciu. Accurate Polyp Segmentation for 3D
   CT Colonography Using Multi-Staged Probabilistic Binary Learning and Compositional Model. CVPR 2008.(pdf)
6. R. Socher, A. Barbu, D. Comaniciu. A Learning Based Hierarchical Model for Vessel Segmentation. 
   IEEE International Symposium on Biomedical Imaging, 2008. (pdf)
7. Y. Zheng, B. Georgescu, A. Barbu, M. Scheuering and D. Comaniciu. Four-Chamber Heart Modeling and
   Automatic Segmentation for 3D Cardiac CT Volumes, SPIE Medical Imaging, 2008.
8. Y. Zheng, A. Barbu, B. Georgescu, M. Scheuering, D. Comaniciu. Fast Automatic Heart Chamber Segmentation
   from 3D CT Data Using Marginal Space Learning and Steerable Features. ICCV 2007 (pdf)
9. S Lakare, M Wolf, L Bogoni, A. Barbu, M Dundar, L Lu, M Salganicoff, Evaluation of a Learning-based
   Component for Suppression of False Positives Located on the Ileo Cecal Valve or Rectal Tube, RSNA 2007
10. A. Barbu, V. Athitsos, B. Georgescu, S. Boehm, P. Durlak, D. Comaniciu. Hierarchical Learning of Curves:
    Application to Guidewire Localization in Fluoroscopy. CVPR 2007 (pdf)
11. S. Lakare, A. Barbu, M. Dundar, M. Wolf, L. Bogoni, D. Comaniciu. Learning-based Component for
    Suppression Rectal Tube False Positives: Evaluation of Performance on 780 CTC Cases, RSNA 2006 (ppt)
12. A. Barbu, L. Bogoni, D. Comaniciu. Hierarchical Part-Based Detection of 3D flexible tubes:Application to
    CT Colonoscopy, MICCAI 2006 (pdf)
13. Z. Tu, X.S. Zhou, A. Barbu, L. Bogoni, D. Comaniciu. Probabilistic 3D Polyp Detection in CT Images: The
    Role of Sample Alignment, CVPR 2006 (pdf)
14. A. Barbu, S.C. Zhu. Incorporating visual knowledge representation in stereo reconstruction,  ICCV 2005 (pdf)
15. A. Barbu, S.C. Zhu. Multigrid and Multi-level Swendsen-Wang Cuts for Hierarchic Graph Partition, CVPR 2004 (pdf)
16. A. Barbu, A.L. Yuille. Motion Estimation by Swendsen-Wang Cuts, CVPR 2004 (pdf)
17. A. Barbu, S.C. Zhu. On the relationship between image and motion segmentation, SCVMA workshop, ECCV 2004 (pdf)
18. A. Barbu, S.C. Zhu. Graph Partition By Swendsen-Wang Cuts, ICCV 2003  (pdf)

Funding:

1. MCS: Research on Detection and Classification of 2D and 3D Shapes in Cluttered Point Clouds. NSF, 400k (CO-PI)
2. Statistical and Semantic Approaches for Object, Activity and Intent Recognition. ONR, 400k (CO-PI)
3. Landmark Detection Using Discriminative Anatomical Network And Active Random Fields. Siemens, 31k (PI) 
4. Cooperative Systems: Task Allocation for Heterogeneous Agent Teams Via Stochastic Clustering Auctions. ARO, 47k (subcontract)

Invited Talks

1. Training an Active Random Field for Real-Time Image Denoising. Max Plank Institute, Saarbrucken, Germany, July 16th, 2008
2. The Swendsen-Wang Cuts Algorithm with Applications in Computer Vision, Georgia Tech University, June 2008
3. Active Random Fields for Real-Time Image Denoising, Siemens Corporate Research, May 2008
4. Hierarchical Image-Motion Segmentation using Swendsen-Wang Cuts, Third Cape Cod MCMC Workshop, Harvard, 2007
5. A General Clustering Sampling Method for Bayesian Inference, Joint Statistical Meetings, Minneapolis, August 10, 2005
6. Swendsen-Wang for Perceptual Grouping. Second Cape Cod Workshop on Monte Carlo Methods, 2004

Patents:

1. Z. Tu, X. Zhou, D. Comaniciu, L. Bogoni, A. Barbu. System and Method for Using Learned Discriminative Models to
   Segment Three dimensional Colon Image Data. Patent No 7,583,831
2. Z. Tu, A. Barbu, D. Comaniciu. Method for Detecting Polyps in a Three Dimensional Image Volume. Patent No 7,558,413

Education:

Research:

1. Active Random Fields - A MRF based model trained together with a fast and suboptimal inference algorithm achieves
   thousands of times speedup without loss in accuracy.
2. Marginal Space Learning - A learning-based optimization method that achieves many orreds of magnitude speedup for
   object detection in large parameter spaces.
3. Graph Partition by Swendsen-Wang Cuts - a stochastic graph partition algorithm.
   It performs fast inference in the graph partition space, guided by a probability model and uses low level cues to speed up
   convergence. The algorithm is ergodic and reversible. 

Software

1. Image Segmentation with Swendsen-Wang Cuts source code.
2. Active Random Fields demo, Berkeley images.