2016

Pairwise Quantization

Babenko, Artem, Arandjelović, Relja, Lempitsky, Victor

Understand

We consider the task of lossy compression of high-dimensional vectors through quantization.

  • We propose the approach that learns quantization parameters by minimizing the distortion of scalar products and squared distances between pairs of points.
  • This is in contrast to previous works that obtain these parameters through the minimization of the reconstruction error of individual points.
  • The proposed approach proceeds by finding a linear transformation of the data that effectively reduces the minimization of the pairwise distortions to the minimization of individual reconstruction errors.

Built on

  • Modeling the shape of the scene: A holistic representation of the spatial envelope

    A. Oliva and A. Torralba · 2001

    Earlier work this paper cites.

  • Distinctive image features from scale-invariant keypoints

    D. G. Lowe · 2004

    Earlier work this paper cites.

  • Imagenet: A large-scale hierarchical image database

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei · 2009

    Earlier work this paper cites.

  • A unified approach to building hybrid recommender systems

    A. Gunawardana and C. Meek · 2009

    Earlier work this paper cites.

  • Approximate nearest neighbor search by residual vector quantization

    Y. Chen, T. Guan, and C. Wang · 2010

    Earlier work this paper cites.

  • Performance of recommender algorithms on top-n recommendation tasks

    P. Cremonesi, Y. Koren, and R. Turrin · 2010

    Earlier work this paper cites.

Similar

  • Product quantization for nearest neighbor search

    H. Jégou, M. Douze, and C. Schmid · 2011

    Cited alongside, same era.

  • High-dimensional signature compression for large-scale image classification

    J. Sánchez and F. Perronnin · 2011

    Cited alongside, same era.

  • Sparse kernel approximations for efficient classification and detection

    A. Vedaldi and A. Zisserman · 2012

    Cited alongside, same era.

  • Optimized product quantization for approximate nearest neighbor search

    T. Ge, K. He, Q. Ke, and J. Sun · 2013

    Cited alongside, same era.

  • Cartesian k-means

    M. Norouzi and D. J. Fleet · 2013

    Cited alongside, same era.

Then

  • Additive quantization for extreme vector compression

    A. Babenko and V. Lempitsky · 2014

    Later among the works it cites.

  • Composite quantization for approximate nearest neighbor search

    T. Zhang, C. Du, and J. Wang · 2014

    Later among the works it cites.

  • Tree quantization for large-scale similarity search and classification

    A. Babenko and V. Lempitsky · 2015

    Later among the works it cites.

  • From captions to visual concepts and back

    H. Fang, S. Gupta, F. N. Iandola, R. K. Srivastava, L. Deng, P. Dollár, J. Gao, X. He, M. Mitchell, J. C. Platt, C. L. Zitnick, and G. Zweig · 2015

    Later among the works it cites.

  • The movielens datasets: History and context

    F. M. Harper and J. A. Konstan · 2016

    Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…