Liefeng Bo, Researcher at Intel Labs

Collaborators: Dieter Fox, Xiaofeng Ren, and Anthony LaMarca

University of Washington Computer Science & Engineering

Box 352350, 101 Paul G. Allen Center for CSE
Seattle, WA 98195-2350

Email: lfb@cs.washington.edu, liefengbo@gmail.com

Biography: I am a research scientist at the Intel Science and Technology Center on Pervasive Computing (ISTC-PC) located in the campus of the University of Washington. My research interests are in Machine Learning, Computer Vision and Robotics. I was a postdoctoral researcher at Computer Science & Engineering at the University of Washington from 2010-2012. From 2008-2009, I worked as a postdoctoral researcher at Toyota Technological Institute at Chicago. From 2002-2007, I was a PhD student at Institute of Intelligent Information Processing in Xidian University, where I received a doctorate in Electronic Engineering, with a specialization in Machine Learning.

 กก

[Research][Publication][Software][CV]

News

  • RGB-Depth Kernel Descriptors are now available!!! Check out the source code and robust RGB-Depth object recognition demos

  • Best Vision Paper Award, ICRA 2011 (Flagship Robotics Conference)

  • 2010 National Excellent Doctoral Dissertation Award (Highest Award for PhD Thesis in China)

  • Strong skills in applying modern machine learning to solve challenging problems

Research Interests

  • Machine Learning: Representation and Feature Learning, Sparse Coding, Support Vector Machines and Kernel Methods, Structured Output Prediction, Graphical Models, Robot Learning

  • Computer Vision and Robotics: Recognition and Detection, RGB-D Vision, RGB-D Kernel Descriptors, Human Pose Estimation, Scene Understanding, Robot Perception

Recent Software (More Software)

Recent Papers (More Papers)

  1. Liefeng Bo, Xiaofeng Ren and Dieter Fox, Unsupervised Feature Learning for RGB-D Based Object Recognition, In International Symposium on Experimental Robotics, (ISER), June 2012. [PDF] [BIB]

  2. Liefeng Bo, Xiaofeng Ren and Dieter Fox, Hierarchical Matching Pursuit for Image Classification: Architecture and Fast Algorithms, Advances in Neural Information Processing Systems (NIPS), December, 2011. [PDF] [BIB][Poster]
    A multi-layer sparse coding model that yields higher accuracy than single-layer sparse coding on top of SIFT
    Batch tree orthogonal matching pursuit = highly efficient sparse coding

  3. Cynthia Matuszek, Nicholas FitzGerald, Liefeng Bo, Luke Zettlemoyer, and Dieter Fox, A joint model of language and perception for grounded attribute learning, In International Conference on Machine Learning (ICML), July 2012.

  4. Xiaofeng Ren, Liefeng Bo and Dieter Fox, RGB-(D) Scene Labeling: Features and Algorithms, In IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), June 2012. [PDF] [BIB] [Code]
    Kernel descriptors + segmentation tree achieves the state-of-the-art results on the NYU and Stanford Background datasets

  5. Kevin Lai, Liefeng Bo, Xiaofeng Ren and Dieter Fox, Detection-based Object Labeling in 3D Scenes, In IEEE International Conference on Robotics and Automation (ICRA), May, 2012. [PDF] [BIB]
    Labeling objects in 3D based on detection score

  6. Liefeng Bo, Xiaofeng Ren and Dieter Fox, Depth Kernel Descriptors for Object Recognition, In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2011. [PDF] [BIB] [Dataset] [Code]
    Kernel Descriptors over depth maps and 3D point clouds

  7. Liefeng Bo, Kevin Lai, Xiaofeng Ren and Dieter Fox, Object Recognition with Hierarchical Kernel Descriptors, In IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), June 2011. [PDF] [BIB] [Dataset] [Code]
    Kernel descriptors over kernel descriptors: a deep architecture

  8. Liefeng Bo, Xiaofeng Ren and Dieter Fox, Kernel Descriptors for Visual Recognition, Advances in Neural Information Processing Systems (NIPS), December, 2010. [PDF] [Spotlight] [Video] [BIB] [Code]
    A general approach to extract local features from pixel attributes that includes popular SIFT and HOG features as special cases
    KDES + EMK + linear SVMs has 77.5% accuracy on Caltech101 and 87.5% accuracy on Scene15 (higher than those reported in the paper)

  9. Kevin Lai, Liefeng Bo, Xiaofeng Ren and Dieter Fox, A Scalable Tree-based Approach for Joint Object and Pose Recognition, In the AAAI Conference on Artificial Intelligence (AAAI), August 2011. [PDF] [BIB] [Dataset] [Code]
    Scalable joint category, instance, and pose recognition system
    Core component of OASIS (Object Aware Situated Interactive System) that was shown live at CES (Consumer Electronics Show) 2011

  10. Kevin Lai, Liefeng Bo, Xiaofeng Ren and Dieter Fox, Sparse Distance Learning for Object Recognition Combining RGB and Depth Information, In IEEE International Conference on Robotics and Automation (ICRA), May, 2011. [PDF] [BIB] [Dataset]
    Best Vision Paper Award

  11. Kevin Lai, Liefeng Bo, Xiaofeng Ren and Dieter Fox, A Large-Scale Hierarchical Multi-View RGB-D Object Dataset, In IEEE International Conference on Robotics and Automation (ICRA), May, 2011. [PDF] [BIB] [Dataset] 
    First Kinect style RGB-Depth object dataset
    300 everyday objects with 250,000 RGB+Depth frames

  12. Liefeng Bo and Cristian Sminchisescu, Twin Gaussian Processes for Structured Prediction, International Journal of Computer Vision (IJCV), vol. 87, pp. 28-52, 2010. [PDF] [BIB] [CODE]
    Learning with structured(multiple) continuous outputs for human pose estimation
    State-of-the-art accuracy on HumanEva

  13. Liefeng Bo and Cristian Sminchisescu, Efficient Match Kernels between Sets of Features for Visual Recognition, Advances in Neural Information Processing Systems (NIPS), December, 2009.  [PDF] [BIB] [Code]
    SIFT + EMK Coding + linear SVM has 74.5% accuracy on Caltech-101 (higher than that reported in the paper)

  14. Jian Peng, Liefeng Bo, and Jinbo Xu, Conditional Neural Fields, Advances in Neural Information Processing Systems (NIPS), December, 2009. [PDF] [BIB] [CODE]
    Integrate multi-layer forward network into conditional random fields