Mutual Information Learned Regressor: an Information-theoretic Viewpoint of Training Regression Systems

no code implementations • 23 Nov 2022 • Jirong Yi, Qiaosheng Zhang, Zhen Chen, Qiao Liu, Wei Shao, Yusen He, Yaohua Wang

We first argue that the MSE minimization approach is equivalent to a conditional entropy learning problem, and then propose a mutual information learning formulation for solving regression problems by using a reparameterization technique.

regression

Deep segmentation networks predict survival of non-small cell lung cancer

1 code implementation • 26 Mar 2019 • Stephen Baek, Yusen He, Bryan G. Allen, John M. Buatti, Brian J. Smith, Ling Tong, Zhiyu Sun, Jia Wu, Maximilian Diehn, Billy W. Loo, Kristin A. Plichta, Steven N. Seyedin, Maggie Gannon, Katherine R. Cabel, Yusung Kim, Xiaodong Wu

Here we show that CNN trained to perform the tumor segmentation task, with no other information than physician contours, identify a rich set of survival-related image features with remarkable prognostic value.

Segmentation Tumor Segmentation

ZerNet: Convolutional Neural Networks on Arbitrary Surfaces via Zernike Local Tangent Space Estimation

1 code implementation • 3 Dec 2018 • Zhiyu Sun, Ethan Rooke, Jerome Charton, Yusen He, Jia Lu, Stephen Baek

In this paper, we propose a novel formulation to extend CNNs to two-dimensional (2D) manifolds using orthogonal basis functions, called Zernike polynomials.

Embedded Spectral Descriptors: Learning the point-wise correspondence metric via Siamese neural networks

no code implementations • 17 Oct 2017 • Zhiyu Sun, Yusen He, Andrey Gritsenko, Amaury Lendasse, Stephen Baek

In this paper, it is proposed a method to improve the similarity metric of spectral descriptors for correspondence matching problems.