Machine Learning for Bioimage Analysis Reading Group 2015

• Jan. 21, 2015 -- Jaydeep DE
  Title: Tracing Filamentary Structures in Neuronal and Retinal Images: a Graph-Theoretical Approach.


• Jan. 28, 2015 -- Chi Xu
  Paper: Piotr Dollar, Peter Welinder and Pietro Perona. Cascaded Pose Regression. CVPR 2010.


• Feb. 04, 2015 -- Dr. Ni Bingbing (guest speaker)
  Title: Recent Advances in Action Recognition (Slides)
  Abstract: Video based Action Recognition has been an important yet very challenging research topic in computer vision during the decades. In this talk, we will give an introduction to the research development in action recognition and will highlight some milestone work. In the meantime, we will introduce several recent emerging works based on middle level visual features for action recognition as well as some new directions.


• Feb. 11, 2015 -- Lin Gu
  Paper: Ehsan Shahrian, Deepu Rajan, Brian Price and Scott Cohen. Improving Image Matting using Comprehensive Sampling Sets . CVPR 2013.


• Feb. 25, 2015 -- Ashwin Nanjappa
  Paper: A. Perez-Escudero, J. Vicente-Page, R. C. Hinz, S. Arganda and G. G. de Polavieja. idTracker: Tracking Individuals in a Group by Automatic Identification of Unmarked Animals . Nature Methods 2014.


• Mar. 04, 2015 -- Xiaowei Zhang
  Paper: Hyunwoo J. Kim, Nagesh Adluru and Maxwell D. Collins et. al. Multivariate General Linear Models (MGLM) on Riemannian Manifolds with Applications to Statistical Analysis of Diffusion Weighted Images . CVPR 2014.


• Mar. 18, 2015 -- Li Cheng
  Paper: O. Freifeld, S. Hauberg and M. J. Black. Model Transport: Towards Scalable Transfer Learning on Manifolds. CVPR 2014.


• Mar. 25, 2015 -- Lakshmi Govindarajan
  Paper: Jia Xu, Alexander G. Schwing and Raquel Urtasun. Tell Me What You See and I will Show You Where It Is. CVPR 2014.


• Apr. 1, 2015 -- Jaydeep DE
  Paper: Engin Turetken, Fethallah Benmansour, Bjoern Andres, Hanspeter Pfister and Pascal Fua. Reconstructing Curvilinear Networks using Path Classifiers and Integer Programming. CVPR 2013.


• Apr. 15, 2015 -- Dr. LIU Li (guest speaker)
  Title: Multilevel Activity Recognition with Probabilistic Interval Model Using Sensors - A Statistical-relational Method.
  Abstract: In pervasive and ubiquitous computing systems, human activity recognition has immense potential in a large number of application domains. In terms of complex activity recognition, most of the current approaches have several limitations. First, most of structure-based approaches such as hidden Markov models and Bayesian networks are limited to model rich temporal and multilevel relationships among activities. Second, it would be difficult for structure-based approaches to build a uniform model that can answer various queries associated with temporal and multilevel relations. Third, semantic-based approaches often lack the expressive power to capture uncertainties associated with their temporal dependencies. Fourth, rules in semantic-based approaches are often manually encoded. To address these issues, we propose a statistical-relational method that combines the Markov logic networks with Allen's temporal relations and two multilevel relations. It is able to answer any temporally and multilevel related queries about the probabilities of activity occurrences through the human readable rules and their weights, which are automatically learned from activity interval temporal patterns.


• May 06, 2015 -- Chi Xu
  Paper: James Steven Supancic III, Gregory Rogez, Yi Yang, Jamie Shotton and Deva Ramanan. Depth-based hand pose estimation: methods, data, and challenges. arXiv:1504.06378.


• May 13, 2015 -- Lin Gu
  Paper: C. Lawrence Zitnick and Piotr Dollar. Edge Boxes: Locating Object Proposals from Edges . ECCV 2014.


• May 20, 2015 -- Ashwin Nanjappa
  Paper: C. Vondrick, A. Khosla, T. Malisiewicz and A. Torralba. HOGgles: Visualizing Object Detection Features. ICCV 2013.


• Jun. 10, 2015 -- Xiaowei Zhang
  Paper 1: Yi Hong, Roland Kwitt, Nikhil Singh, Brad Davis, Nuno Vasconcelos and Marc Niethammer. Geodesic Regression on the Grassmannian. ECCV 2014.
  Paper 2: Yi Hong, Nikhil Singh, Roland Kwitt, Nuno Vasconcelos and Marc Niethammer. Parametric Regression on the Grassmannian. arXiv:1505.03832.


• Jun. 17, 2015 -- Li Cheng
  Paper: Q. Xie, S. Kurtek, H. Le and A. Srivastava. Parallel Transport of Deformations in Shape Space of Elastic Surfaces. ICCV 2013.


• Jun. 24, 2015 -- Lakshmi Govindarajan
  Paper: E. Eyjolfsdottir, S. Branson, X. Burgos-Artizzu, E. Hoopfer, J. Schor, D. Anderson and P. Perona. Detecting Social Actions of Fruit Flies. ECCV 2014.


• Jul. 22, 2015 -- Dr. Zhang Yu
  Paper: Yu Zhang, Jianxin Wu and Jianfei Cai. Compact Representation for Image Classification: To Choose or to Compress? CVPR 2014.
  Abstract: In large scale image classification, features such as Fisher vector or VLAD have achieved state-of-the-art results. However, the combination of large number of examples and high dimensional vectors necessitates dimensionality reduction, in order to reduce its storage and CPU costs to a reasonable range. In spite of the popularity of various feature compression methods, this paper argues that feature selection is a better choice than feature compression. We show that strong multicollinearity among feature dimensions may not exist, which undermines feature compression's effectiveness and renders feature selection a natural choice. We also show that many dimensions are noise and throwing them away is helpful for classification. We propose a supervised mutual information (MI) based importance sorting algorithm to choose features. Combining with 1-bit quantization, MI feature selection has achieved both higher accuracy and less computational cost than feature compression methods such as product quantization and BPBC.


• Jul. 29, 2015 -- Shijie Xiao (guest speaker)
  Title: Low-Rank Representation: Optimization and Variants
  Abstract: Initially proposed for dealing with the problem of subspace clustering, Low-Rank Representation (LRR) has shown promising results on various applications such as motion segmentation and face clustering. Whereas, most of existing LRR solvers directly operate on its formulation containing the original data matrix, thus making the optimization inefficient. In this talk, we will first introduce an efficient LRR solver which is based on reformulating LRR as a new optimization problem. This solver achieves order-of-magnitude speedup over several existing solvers, and it can be readily incorporated into a distributed framework for further acceleration. Moreover, we will introduce two variants of LRR for 1) face clustering in videos and 2) face naming with weakly labeled images, respectively. Experiments on several real-world datasets demonstrate the effectiveness of these two variants of LRR.


• Aug. 12, 2015 -- Yongzhong Yang
  Paper: Xiao Sun, Yichen Wei, Shuang Liang, Xiaoou Tang and Jian Sun. Cascaded Hand Pose Regression. CVPR 2015.


• Aug. 20, 2015 -- Chi Xu
  Paper: Luis Ferraz, Xavier Binefa and Francesc Moreno-Noguer. Very Fast Solution to the PnP Problem with Algebraic Outlier Rejection. CVPR 2014.


• Aug. 25, 2015 -- Lin Gu
  Paper:


• Sep. 3, 2015 -- Ashwin Nanjappa
  Paper: Jumpei Matsumoto, Susumu Urakawa, Yusaku Takamura et al. A 3D-Video-Based Computerized Analysis of Social and Sexual Interactions in Rats. PLOS ONE 2013.


• Sep. 17, 2015 -- Xiaowei Zhang
  Paper: Marko Ristin, Matthieu Guillaumin, Juergen Gall, and Luc Van Gool. Incremental Learning of Random Forests for Large-Scale Image Classification. IEEE TPAMI 2015.


• Oct. 08, 2015 -- Xiaowei Zhang
  Title: An Overview of the Machine Learning Summer School 2015 Kyoto


• Oct. 22, 2015 -- Li Cheng
  Paper: A. Sironi, E. Turetken, V. Lepetit and P. Fua. Multiscale Centerline Detection. IEEE TPAMI, in press, 2015


• Nov. 19, 2015 -- Lakshmi Govindarajan
  Paper: V. Ramakrishna, D. Munoz, M. Hebert, J. Andrew Bagnell, and Y. Sheikh. Pose Machines: Articulated Pose estimation via Inference Machines. ECCV 2014.


• Nov. 26, 2015 -- Chenglong Bao (guest speaker)
  Title: Dictionary learning for sparse coding: Algorithms and convergence analysis.