Search

In the past decades, a variety of learning-based algorithms have been emerged to try to explore a better solution for stereo matching by leveraging various machine learning algorithms. For enriching learning-based stereo matching algorithm's methodologies, we cast the disparity estimation as a regression problem by leveraging Structured Support Vector Machine (SSVM) in this paper. There are three categories of features have been extracted on account of disparity cues for training the SSVM. Particularly, one of the three feature is named as 'Coarse-to-Fine PatchMatch Filtering', which effectively exploits region and pixel disparity cues. For attaining region disparity cues, we adopt MeshStereo and MeshStereo with Cross-Scale algorithms; for attaining pixel disparity cues, PatchMatch and Cross-Scale PatchMatch stereo matching algorithms are utilized. Performance evaluations on Middlebury v.2 and v.3 stereo data sets demonstrate that the proposed algorithm reveals comparable accuracy with other challenging learning-based ones. It is worth pointing out that our proposal performs over several orders of magnitude faster than others on training time. [ABSTRACT FROM AUTHOR]

Human emotion recognition is a technique for identifying human emotions with respect to various aspects of human life, such as in decision-making, detecting lies, assessing social behaviour, measuring brain-related activity and identifying the personality of a person. In general, this paper aims to propose an efficient EEG-based cognitive personality trait detection based on the human emotion recognition system. Two different algorithms were proposed for this system, namely the normalised window-short time Fourier transform (NW-STFT) algorithm and the cognitive mapping-based Hebbian learning (CM-HL) algorithm. The NW-STFT algorithm dynamically fixes the window size for feature extraction using EEG-based emotion recognition, and the CM-HL algorithm is proposed for mapping the top five big personality traits with the help of recognised primary emotions. The pre-processing method reduces the contaminated artefacts from the benchmark dataset, as well as the brain signal's band power frequency range analysis time. The proposed emotion recognition method has shown a significant improvement in accuracy, leading to greater enhancement in personality-trait mapping. The outcome shows that the proposed two algorithms are effective in finding emotional and personality trait mapping. query Please check the edit made in the article title. [ABSTRACT FROM AUTHOR]

Sign language is a natural language mostly used by persons with hearing- and speech-based impairments to communicate with other people. In these modern times, sign language guides are used to eliminate the communication gap between people having hearing impairments and those with or without hearing impairments; however, they are very limited in number. To solve this challenge, automatic sign language recognition systems are developed to better reduce the communication gap for people with hearing disabilities. This paper presents the development of an automatic Amharic sign language translator which translates Amharic alphabet signs into their corresponding text using digital image processing and machine learning algorithms. The proposed system has four major developmental stages which include preprocessing, segmentation, feature extraction and classification. A total number of thirty-four features were extracted from shape, motion and color of hand gestures to represent both the base and derived class of Amharic sign characters. Classification models were built using artificial neural network (ANN) and multi-class support vector machine (SVM). The results show that the recognition system is capable of recognizing the Amharic alphabet signs with an average accuracy of 80.82% and 98.06% using the ANN and SVM classifiers, respectively. [ABSTRACT FROM AUTHOR]

In the task of semantic segmentation, researchers often use self-attention module to capture long-range contextual information. These methods are often effective. However, the use of the self-attention module will cause a problem that cannot be ignored, that is, the huge consumption of computing resources. Therefore, how to reduce the resource consumption of the self-attention module under the premise of ensuring performance is a very meaningful research topic. In this paper, we propose a Sparse Attention Model combined with a powerful multi-task feature extraction network for semantic segmentation. Compared with the classic self-attention model, our Sparse Attention Model does not calculate the inner product between pairs of all vectors. Instead, we first sparse the feature block Query and the feature block Key defined in self-attention module through the credit matrix generated by the pre-output. Then, we perform similarity modeling on the two sparse feature blocks. Meanwhile, to ensure that the vectors in Query could capture dense contextual information, we design a Class Attention Module and embed it into Sparse Attention Module. Note that, compared with Dual Attention Network for scene segmentation, our attention module greatly reduces the consumption of computing resources while ensuring the accuracy. Furthermore, in the stage of feature extraction, the use of downsampling will cause serious loss of detailed information and affect the segmentation performance of the network, so we adopt a multi-task feature extraction network. It learns semantic features and edge features in parallel, and we feed the learned edge features into the deep layer of the network to help restore detailed information for capturing high-quality semantic features. We do not use pure concatenation. Instead, we extract the edge features related to each channel by element-wise multiplication before concatenation. Finally, we conduct experiments on three datasets: Cityscapes, PASCAL VOC2012 and ADE20K, and obtain competitive results. [ABSTRACT FROM AUTHOR]

Classifying texture images relies heavily on the quality of the extracted features. However, producing a reliable set of features is a difficult task that often requires human intervention to select a set of prominent primitives. The process becomes more difficult when it comes to fuse low-level descriptors because of data redundancy and high dimensionality. To overcome these challenges, several approaches use machine learning to automate primitive detection and feature extraction while combining low-level descriptors. Nevertheless, most of these approaches performed the two processes separately while ignoring the correlation between them. In this paper, we propose a genetic programming (GP)-based method that combines the two well-known features of histograms of oriented gradients and local binary patterns. Indeed, a three-layer tree-based binary program is learned using genetic programming for each pair of classes. The three layers incorporate patch detection, feature fusion and classification in the GP optimization process. The feature fusion function is designed to handle different variations, notably illumination and rotation, while reducing dimensionality. The proposed method has been compared, using six challenging collections of images, with multiple domain-expert GP and non-GP methods for binary and multi-class classifications. Results show that the proposed method significantly outperforms or achieves similar performance to relevant methods from the state-of-the-art, even with a limited number of training instances. [ABSTRACT FROM AUTHOR]

Current research on deep learning algorithms focuses on the reconstruction of hyperspectral images from a single RGB image. However, this does not consider the feature information between regions, so the feature capture of context is insufficient. This causes the quality of reconstructed hyperspectral images to be low. We propose correcting this with a fusiform multi-scale pixel self-attention (FMPSA) network. The proposed FMPSA consists of a fusiform multi-scale feature extraction (FMFE) module cascaded with several multi-scale adaptive residual attention blocks (MARABs). FMFE extracts multi-scale detail features by interleaving dual components to avoid degrading spectral reconstruction quality due to local and edge spatial information loss. Each MARAB consists of paired FMFE-Left and FMFE-Right components, an optimal non-local model, a pixel self-attention module, a LayerNorm layer, a multilayer perceptron with Gelu nonlinearity, and long-short dual residual connection, which can be regarded as a residual structure based on a pixel self-attention mechanism. MARAB can adaptively track regions containing feature-rich information for more accurate hyperspectral reconstruction with a hierarchical focus on the salient pixels. The proposed FMPSA was applied to the NTIRE 2020 hyperspectral dataset. Experimental results show that the proposed method outperforms current methods in terms of MRAE and RMSE. [ABSTRACT FROM AUTHOR]

Image analysis is an essential field for several topics in the life sciences, such as biology or botany. In particular, the analysis of seeds (e.g. fossil research) can provide important information on their evolution, the history of agriculture, plant domestication, and diets knowledge in ancient times. This work presents software that performs image analysis for feature extraction and classification from images containing seeds through a novel and unique framework. In detail, we propose two plugins for ImageJ, one able to extract morphological, texture, and colour features from seed images, and another to classify seeds using the extracted features. The experimental results demonstrated the correctness and validity of both the extracted features and the classification predictions on two public seeds datasets, showing that combining the handcrafted features with the Random Forest classifier can reach outstanding performance on both datasets. The proposed tool is easily extendable to other fields of image analysis. [ABSTRACT FROM AUTHOR]

Summarization is important in sports video analysis; it gives a more compact and interesting representation of content. The automatic cricket video summarization is more challenging as it contains several rules and longer match duration. In this research, a hybrid machine learning approach is proposed to summarize cricket video. It analyzes the excitement, object, and event-based features for the detection of key events from the cricket video. First, the audio is analyzed for the extraction of the exciting clips by using an adaptive threshold, speech-to-text framework, and Stacked Gated Recurrent Neural Network with Attention Module (SGRNN-AM). Then, the scenes of each exciting clip are classified with a new Hybrid Rotation Forest Deep Belief Network (HRF-DBN). Next, the characters and action features are extracted from the scorecard region of each key frame and umpire frames of exciting clips. Finally, SGRNN-AM model is used to detect key events including fours, sixes, and wickets. The accuracy of the proposed SGRNN-AM video summarization model is increased with an attention module in the hidden outputs of Gated Recurrent Unit (GRU) for selecting the significant features. The performance of the suggested technique has been improved on various collections of cricket videos. It achieved a precision of 96.82 % and an accuracy of 96.32 % that proves its effectiveness. [ABSTRACT FROM AUTHOR]