Posts by 'leohung'

"Semi-Supervised Hashing for Scalable Image Retrieval"

J. Wang et al, "Semi-Supervised Hashing for Scalable Image Retrieval," CVPR, 2010.

The following figures and formula are all copied from this paper.

Novelty

Unsupervised learning method, like LSH, would ignore the relationship of labels; however, supervised learning method modified from LSH, like RBM or SH, would cost time to ...

"Mairal et al. Online dictionary learning for sparse coding"

Mairal et al. Online dictionary learning for sparse coding. ICML 2009.

The following figures are all from this paper.

The paper goal is to propose a online learning method to learn sparse coding dictionary, the basis set which can represent specific data through linear combination, for large scale data.

Algorithm ...

"Aggregating local descriptors into a compact image representation"

Herve Jegou, et al., Aggregating local descriptors into a compact image representation, Proc. IEEE CVPR'10 The goal of the work is reducing the computing time and the memory usage without lossing too much accuracy in image retrieval for large scale data. This work consists of two parts: VLAD and ...

"Efficient visual search of videos cast as text retrieval"

"Efficient visual search of videos cast as text retrieval," J. Sivic, and A. Zisserman, IEEE TPAMI, 2009 Paper Authors: Josef Sivic and Andrew Zisserman The following figures are all from this paper.

outline

J. Sivic, and A. Zisserman proposed this method which analogizes searching in video to in text corpora: representing ...

"Distinctive Image Features from Scale-Invariant Keypoints"

"Distinctive Image Features from Scale-Invariant Keypoints," Lowe, IJCV, 2004

Paper author: David G. Lowe The following figures are all from this paper.

Novelties, contributions, assumption

This work introduce a set of promising affine invariant features, which can be used to identified specific object varied with background, illumination, rotation degree.

Questions ...

"PLSI & LDA"

PLSI & LDA

Origin source: "Probabilistic latent semantic indexing," T. Hofmann, SIGIR, 1999. "Latent Dirichlet allocation," D. Blei, A. Ng, and M. Jordan. . Journal of Machine Learning Research, 3:993–1022, January 2003 The following figures and formula are copied from above papers.

PLSI and LDA are both hidden-class decomposition method ...

"Nonlinear dimensionality reduction by locally linear embedding"

"Nonlinear dimensionality reduction by locally linear embedding," Roweis & Saul, Science, 2000.

The following figures are all copied from the essay.

nonlinear-dimension-reduction

The main purpose of this method is to reduce the data from high dimensions to low dimensions while maintaining instances local relative distances at the same time. There are three ...

"Support vector learning for ordinal regression"

"Support vector learning for ordinal regression," R. Herbrich, T. Graepel, K. Obermayer, ICANN, 1999

Authors: R. Herbrich, T. Graepel, K. Obermayer The following figures are all copies from the paper.

The classical supervised machine learning methods deal with classification problem and regression problem. Instead, this research proposed a method to ...

"Learning to rank: from pairwise approach to listwise approach"

"Learning to rank: from pairwise approach to listwise approach," Cao, ICML, 2007.

Authors: Zhe Cao, Tao Qin, Tie-Yan Liu, Ming-Feng Tsai, Hang Li The following figures are all copied from this paper.

Algorithm

Unlike pairwise rank-learning method like RankSVM, this work proposed a listwise rank learning framework where training instances ...

"A Global Geometric Framework for Nonlinear Dimensionality Reduction"

Authors: Joshua B. Tenenbaum, Vin de Silva, John C. Langford

The following figures are copied from the original paper.

This work introduces a method called Isomap for dimension reduction specifically on non-linear data. The algorithm composes of three steps:

(1) Find neighbors for each points in input space. To do ...