Assistive Technology based Solution for Hearing Impairment using Smartphone

Developing a system for sign language recognition becomes essential for the deaf as well as a mute person. The recognition system acts as a translator between a disabled and an able person. This eliminates the hindrances in the exchange of ideas. Most of the existing systems are very poorly designed with limited support for the needs of their day to day facilities. The proposed system embedded with gesture recognition capability has been introduced here which extracts signs from a video sequence and displays them on screen. On the other hand, a speech to text as well as text to speech system is also introduced to further facilitate the grieved people. To get the best out of a human-computer relationship, the proposed solution consists of various cutting-edge technologies and Machine Learning based sign recognition models that have been trained by using TensorFlow and Keras library. The proposed architecture works better than several gesture recognition techniques like background elimination and conversion to HSV

An Adaptive Genetic Algorithm-based Background Elimination Model for English Text

In this paper, an adaptive genetic algorithm is used to conduct an in-depth study and analysis of English text background elimination, and a corresponding model is designed. The curve results after the initial character editorialization are curved and transformed, and the adaptive genetic algorithm is used for the transformation to solve the influence of multiple inflection points of curve images on feature extraction. Then, using the minimum deviation method, the error values of the input characters and the sample set in the spatial coordinate system are calculated, and the deviation values of the angle and the straight line are used to match the characters with the smallest deviation value to match the highest degree. A genetic algorithm is introduced to iterate the feature sets of angles and line segments, and the optimal features are finally derived in the process of cross evolution of generations to improve the recognition accuracy. And the character library is used as input items for average grouping for experiments, and the obtained feature sets are put into the position matrix and compared with the samples in the database one by one. It is found that the improved stroke-structure feature extraction algorithm based on a genetic algorithm can improve the recognition accuracy and better accomplish the recognition task with better results compared to others. Finally, by analyzing the limitations and characteristics of traditional particle swarm optimization algorithm and differential evolution algorithm, and giving full play to the advantages and applicability of different algorithms, a new differential evolution particle swarm algorithm with better performance and more stable performance is proposed. The algorithm is based on the PSO algorithm, and when the population update of the PSO algorithm is stagnant and the search space is limited, the crossover and mutation operations of the DE algorithm are used to perturb the population, increase the diversity of the population, and improve the global optimization ability of the algorithm. The algorithm is tested on a common dataset for text mining to verify the effectiveness and feasibility of the algorithm.

CaBagE: A Cas9-based Background Elimination strategy for targeted, long-read DNA sequencing

A substantial fraction of the human genome is difficult to interrogate with short-read DNA sequencing technologies due to paralogy, complex haplotype structures, or tandem repeats. Long-read sequencing technologies, such as Oxford Nanopore’s MinION, enable direct measurement of complex loci without introducing many of the biases inherent to short-read methods, though they suffer from relatively lower throughput. This limitation has motivated recent efforts to develop amplification-free strategies to target and enrich loci of interest for subsequent sequencing with long reads. Here, we present CaBagE, a method for target enrichment that is efficient and useful for sequencing large, structurally complex targets. The CaBagE method leverages the stable binding of Cas9 to its DNA target to protect desired fragments from digestion with exonuclease. Enriched DNA fragments are then sequenced with Oxford Nanopore’s MinION long-read sequencing technology. Enrichment with CaBagE resulted in a median of 116X coverage (range 39–416) of target loci when tested on five genomic targets ranging from 4-20kb in length using healthy donor DNA. Four cancer gene targets were enriched in a single reaction and multiplexed on a single MinION flow cell. We further demonstrate the utility of CaBagE in two ALS patients with C9orf72 short tandem repeat expansions to produce genotype estimates commensurate with genotypes derived from repeat-primed PCR for each individual. With CaBagE there is a physical enrichment of on-target DNA in a given sample prior to sequencing. This feature allows adaptability across sequencing platforms and potential use as an enrichment strategy for applications beyond sequencing. CaBagE is a rapid enrichment method that can illuminate regions of the ‘hidden genome’ underlying human disease.

Technological Aids for Deaf and Mute in Modern World

Background:: People suffering from hearing and speaking disabilities have a few ways of communicating with other people. One of these is to communicate through the use of sign language. Objective:: Developing a system for sign language recognition becomes essential for deaf as well as a mute person. The recognition system acts as a translator between a disabled and an able person. This eliminates the hindrances in exchange of ideas. Most of the existing systems are very poorly designed with limited support for the needs of their day to day facilities. Methods:: The proposed system embedded with gesture recognition capability has been introduced here which extracts signs from a video sequence and displays them on screen. On the other hand, a speech to text as well as text to speech system is also introduced to further facilitate the grieved people. To get the best out of human computer relationship, the proposed solution consists of various cutting-edge technologies and Machine Learning based sign recognition models which have been trained by using Tensor Flow and Keras library. Result:: The proposed architecture works better than several gesture recognition techniques like background elimination and conversion to HSV because of sharply defined image provided to the model for classification. The results of testing indicate reliable recognition systems with high accuracy that includes most of the essential and necessary features for any deaf and dumb person in his/her day to day tasks. Conclusion:: It’s the need of current technological advances to develop reliable solutions which can be deployed to assist deaf and dumb people to adjust to normal life. Instead of focusing on a standalone technology, a plethora of them have been introduced in this proposed work. Proposed Sign Recognition System is based on feature extraction and classification. The trained model helps in identification of different gestures.

Deep Fuzzy Multi-Object Categorization in Scene

Object Categorization is the process of, identifying and labelling the various distinct Classes (Categories), in the given input image. The Deep Fuzzy Multi-Object Categorization (DFMOC) model, combines the learning capability of Convolution Neural Networks (CNN) and the uncertainty-managing ability of Fuzzy system, for carrying out the categorization task. This work starts with Background Elimination process for ensuring the image clarity, followed by Fuzzification and Fuzzy Entropy computation. Simple fuzzy sets are to be framed, by employing Fuzzy C-Means (FCM) algorithm, for fuzzification of the input image. Thresholding Block is incorporated, for determining the clusters . The Fuzzy Entropy Computation (FEC) is done, to minimize the Fuzziness rate of the acquired input and consequently, the layers of CNN are trained in accordance with that. Caltech-101 Dataset is been utilized for analysis. Average Precision Rate of Categorization (APRC), along with other metrics namely Time taken and Error Rate, shows that DFMOC model performs better than other models

CaBagE: a Cas9-based Background Elimination strategy for targeted, long-read DNA sequencing

A substantial fraction of the human genome is difficult to interrogate with short-read DNA sequencing technologies due to paralogy, complex haplotype structures, or tandem repeats. Long-read sequencing technologies, such as Oxford Nanopore’s MinION, enable direct measurement of complex loci without introducing many of the biases inherent to short-read methods, though they suffer from relatively lower throughput. This limitation has motivated recent efforts to develop amplification-free strategies to target and enrich loci of interest for subsequent sequencing with long reads. Here, we present CaBagE, a novel method for target enrichment that is efficient and useful for sequencing large, structurally complex targets. The CaBagE method leverages the stable binding of Cas9 to its DNA target to protect desired fragments from digestion with exonuclease. Enriched DNA fragments are then sequenced with Oxford Nanopore’s MinION long-read sequencing technology. Enrichment with CaBagE resulted in up to 416X coverage of target loci when tested on five genomic targets ranging from 4-20kb in length using healthy donor DNA. Four cancer gene targets were enriched in a single reaction and multiplexed on a single MinION flow cell. We further demonstrate the utility of CaBagE in two ALS patients with C9orf72 short tandem repeat expansions to produce genotype estimates commensurate with genotypes derived from repeat-primed PCR for each individual. With CaBagE there is a physical enrichment of on-target DNA in a given sample prior to sequencing. This feature allows adaptability across sequencing platforms and potential use as an enrichment strategy for applications beyond sequencing. CaBagE is a rapid enrichment method that can illuminate regions of the ‘hidden genome’ underlying human disease.

Multiple Constrained Reweighted Penalized Least Squares for Spectral Baseline Correction

Baseline drift occurs in various measured spectra, and the existence of a baseline signal will influence qualitative and quantitative analyses. Therefore, it is necessary to perform baseline correction or background elimination before spectral analysis. In this paper, a multiple constrained asymmetric least squares method based on the penalized least squares principle is proposed for baseline correction. The method takes both baseline and peak characteristics into account. Based on the prior knowledge that the left and right boundaries of characteristic peaks should be symmetrical, additional constraints of penalized least squares are added, which ensure the symmetry of spectra. The experimental results of the proposed method on simulated spectra are compared with existing baseline correction methods to verify the accuracy and adaptability of the proposed method. The method is also successfully applied to the baseline correction of real spectra. The results show that it can be effective for estimating the baseline. In addition, this method can also be applied to the baseline correction of other similar spectral signals.

Data Augmentation Method by Applying Color Perturbation of Inverse PSNR and Geometric Transformations for Object Recognition Based on Deep Learning

Deep learning is applied in various manufacturing domains. To train a deep learning network, we must collect a sufficient amount of training data. However, it is difficult to collect image datasets required to train the networks to perform object recognition, especially because target items that are to be classified are generally excluded from existing databases, and the manual collection of images poses certain limitations. Therefore, to overcome the data deficiency that is present in many domains including manufacturing, we propose a method of generating new training images via image pre-processing steps, background elimination, target extraction while maintaining the ratio of the object size in the original image, color perturbation considering the predefined similarity between the original and generated images, geometric transformations, and transfer learning. Specifically, to demonstrate color perturbation and geometric transformations, we compare and analyze the experiments of each color space and each geometric transformation. The experimental results show that the proposed method can effectively augment the original data, correctly classify similar items, and improve the image classification accuracy. In addition, it also demonstrates that the effective data augmentation method is crucial when the amount of training data is small.