Constraint-Based, Minimum Description Length Computation and Total Sampling for Efficient Constraint Problems – Non-parametric sparse coding (NSCC) is an efficient sparse coding algorithm for sparse coding which has been extensively studied in the literature. Although NSCC works well for many real-world problems, its simplicity and high computational complexity makes it difficult to learn the code to solve these problems. In this paper, we demonstrate that NSCC, using a sparse coding algorithm, can be solved to the best of our knowledge without any sparsity and by a single sparse coding algorithm in two steps of learning. Moreover, we prove that the problem of learning a sparse coding algorithm to solve non-parametric sparse coding is NP-hard. The results show the effectiveness of NSCC, and we hope that this has not hampered the other methods to solve non-parametric sparse coding.

A new technique called negative image enhancement (NNE) has been proposed to exploit image attributes such as background, background color and foreground in a way that can increase the quality of a visual scene. However, only a limited amount of training data is available for the NNE approach. This paper proposes a novel approach based on the use of the image dimensionality score to enhance the quality of the image in a deep learning framework. We show that our proposed technique can effectively enhance the image in the same way as the image dimensionality score would enhance. The evaluation on several popular image enhancement benchmarks shows that our proposed method significantly improves performance compared to other similar approaches.

Inverted Reservoir Computing

A New Algorithm for Unsupervised Learning of Motion Patterns from Moving Object Data

Constraint-Based, Minimum Description Length Computation and Total Sampling for Efficient Constraint Problems

Bayesian Deep Learning for Deep Reinforcement Learning

Learning from Negative News by Substituting Negative Images with Word2vecA new technique called negative image enhancement (NNE) has been proposed to exploit image attributes such as background, background color and foreground in a way that can increase the quality of a visual scene. However, only a limited amount of training data is available for the NNE approach. This paper proposes a novel approach based on the use of the image dimensionality score to enhance the quality of the image in a deep learning framework. We show that our proposed technique can effectively enhance the image in the same way as the image dimensionality score would enhance. The evaluation on several popular image enhancement benchmarks shows that our proposed method significantly improves performance compared to other similar approaches.