Improved Deep Hashing with Scalable Interblock for Tourist Image Retrieval

Tourist image retrieval has attracted increasing attention from researchers. Mainly, supervised deep hash methods have significantly boosted the retrieval performance, which takes hand-crafted features as inputs and maps the high-dimensional binary feature vector to reduce feature-searching complexity. However, their performance depends on the supervised labels, but few labeled temporal and discriminative information is available in tourist images. This paper proposes an improved deep hash to learn enhanced hash codes for tourist image retrieval. It jointly determines image representations and hash functions with deep neural networks and simultaneously enhances the discriminative capability of tourist image hash codes with refined semantics of the accompanying relationship. Furthermore, we have tuned the CNN to implement end-to-end training hash mapping, calculating the semantic distance between two samples of the obtained binary codes. Experiments on various datasets demonstrate the superiority of the proposed approach compared to state-of-the-art shallow and deep hashing techniques.

Application of binary dynamical systems in the problem of classification of Boolean vectors

The article proposes a method based on using binary dynamical systems in the classification problem for Boolean vectors (binary feature vectors). This problem has practical application in various fields of science and industry, for example, bioinformatics, remote sensing of natural objects, smart devices of the Internet of things, etc. Binary synchronous autonomous nonlinear dynamic models with an unknown characteristic matrix are considered. Matrix elements are chosen in such a way that the Boolean reference vectors are equilibrium states of the binary dynamic model. The attraction regions of equilibrium states act as classes (one reference vector corresponds to each class). The classified vector is the initial state of the model. Simple and aggregated classifiers are considered. The proposed method is demonstrated using an illustrative example.

Uncovering feature interdependencies in high-noise environments with stepwise lookahead decision forests

Conventionally, random forests are built from “greedy” decision trees which each consider only one split at a time during their construction. The sub-optimality of greedy implementation has been well-known, yet mainstream adoption of more sophisticated tree building algorithms has been lacking. We examine under what circumstances an implementation of less greedy decision trees actually yields outperformance. To this end, a “stepwise lookahead” variation of the random forest algorithm is presented for its ability to better uncover binary feature interdependencies. In contrast to the greedy approach, the decision trees included in this random forest algorithm, each simultaneously consider three split nodes in tiers of depth two. It is demonstrated on synthetic data and financial price time series that the lookahead version significantly outperforms the greedy one when (a) certain non-linear relationships between feature-pairs are present and (b) if the signal-to-noise ratio is particularly low. A long-short trading strategy for copper futures is then backtested by training both greedy and stepwise lookahead random forests to predict the signs of daily price returns. The resulting superior performance of the lookahead algorithm is at least partially explained by the presence of “XOR-like” relationships between long-term and short-term technical indicators. More generally, across all examined datasets, when no such relationships between features are present, performance across random forests is similar. Given its enhanced ability to understand the feature-interdependencies present in complex systems, this lookahead variation is a useful extension to the toolkit of data scientists, in particular for financial machine learning, where conditions (a) and (b) are typically met.

A Novel HDR Image Zero-Watermarking Based on Shift-Invariant Shearlet Transform

In this paper, a novel high dynamic range (HDR) image zero-watermarking algorithm against the tone mapping attack is proposed. In order to extract stable and invariant features for robust zero-watermarking, the shift-invariant shearlet transform (SIST) is used to transform the HDR image. Firstly, the HDR image is converted to CIELAB color space, and the L component is selected to perform SIST for obtaining the low-frequency subband containing the robust structure information of the image. Secondly, the low-frequency subband is divided into nonoverlapping blocks, which are transformed by using discrete cosine transform (DCT) and singular value decomposition (SVD) to obtain the maximum singular values for constructing a binary feature image. To increase the watermarking security, a hybrid chaotic mapping (HCM) is employed to get the scrambled watermark. Finally, an exclusive-or operation is performed between the binary feature image and the scrambled watermark to compute robust zero-watermark. Experimental results show that the proposed algorithm has a good capability of resisting tone mapping and other image processing attacks.

Improvement of BNSGAII to Increase the Accuracy and Reduction of Feature Selection Subsets

Feature selection is a process for the elimination of irrelevant and redundant features from a dataset in order to improve learning performance in terms of accuracy and time to build a model from the subsets. The conventional techniques in this regard have limitations such as the high computational overhead for training, even in moderate datasets. Although attention has been paid to the development of rapid and accurate detection techniques, finding a dataset of features that could increase detection accuracy is paramount. The issue with feature selection is the NP-hard problem; therefore, an optimal solution cannot be guaranteed. The present study aimed to propose a new solution for the non-dominated sorting genetic algorithm (NSGA II) by making it binary through the Sigmoid transfer function and a thresholding device for binary feature selection in order to improve the performance in feature selection problems in terms of the accuracy and reduction of the subset dimensions. In addition, the efficiency of the proposed algorithm in reducing the mentioned parameters was measured through comparison with other methods in the four datasets of breast cancer, hepatitis, heart, and diabetes.

Visible verbal morphology: Morpheme constancy in Germanic and Romance verbal inflection

In different spelling systems, different grades of morpheme constancy can be found: German has a high degree of morpheme constancy (especially stem constancy, for example rennen – rennt both forms with <nn>), while English has comparatively less (running – run, only the disyllabic form with <nn>). This paper investigates the interaction between stems and verbal inflectional suffixes in terms of constancy in three Germanic languages (Dutch, English, German) and five Romance languages (French, Italian, Portuguese, Romanian, Spanish). Verbal inflection is always the most widespread inflection, so it is a well-defined area for getting an idea of how spelling systems may function. For the Germanic languages, this analysis will primarily focus on the alternation between monosyllabic and disyllabic forms. For the Romance languages, it will focus on the <c>/ <g>-alternations in interaction with the following vowel. The aim is to describe a scale of morphological spelling: The alternation of <c> and <ç> is not an instance of constancy, but of similarity, something between constancy and non-constancy. Morpheme constancy is no longer a binary feature. Comparing verbal inflection takes us another step towards the development of typological parameters for visible morphology.