Affect Recognition Accuracy Research Articles

Semi-supervised action recognition with dynamic temporal information fusion

The most advanced semi-supervised models available are based on images for innovation, and the use of semi-supervised learning models augmented with temporal data for video-level action recognition still suffers from severe model mismatches, and the models are not sufficiently capable of capturing both local and global information about the action. Secondly the use of constant-threshold pseudo-labeling leads to low utilization of unlabeled data for difficult actions in the early stages of training, poor pseudo-labeling quality and affects recognition accuracy. To make the semi-supervised framework FixMatch more suitable for action recognition, we propose Time-Mixer and Dynamic Threshold, respectively. Time-Mixer explores complementary information between time sequences through the fusion of two-channel temporal context information. Dynamic Threshold utilizes a new core mapping function (Normal Distribution Function) to enhance pseudo-labeling quality. Extensive experiments were conducted on three action recognition datasets (Kinetics-400, UCF-101, and HMDB-51). Comprehensive experiments show that the performance of the semi-supervised model in action recognition improves considerably after using dynamic thresholding and temporal context information fusion, with a 14.4% improvement over the baseline and a 1.8% improvement over the TG (with a labeling rate of 10%) in UCF101, whereas an overall good performance is obtained for DTIF.

Multiple receiver specific emitter identification

AbstractSpecific emitter identification (SEI) is a technique for identifying emitters based on the principle that the hardware chain is not ideal, causing the emitted signal to contain emitter‐specific information. However, the receiver is also non‐ideal, which affects recognition accuracy and introduces receiver‐specific information that makes SEI difficult to generalise across receiving systems. In this work, a new multi‐receiver receiving and processing system (MR‐SEI) scheme is proposed to mitigate the influence of receivers based on the analysis of receiver distortion models. After receiving and processing in a specific manner, recognition performance can be enhanced. Therefore, extracted features can be shared among different receivers and platforms, and can even be applied to newly added receivers. The concept of common waveform (CW) is first defined, referring to the received signal without receiver distortions. Different receiving devices are working synchronously, and the CW is estimated using multiple copies of the signal obtained from multiple receivers through the iterative reweighted least squares (IRLS) method. For each receiver, a maximum linear correlation algorithm is proposed to calculate the received signal without being affected by distortions. Experimental results show that the proposed scheme can enhance identification performance. With the increase in the number of receivers, the improvement is more noticeable. Using 10 distorted receivers operating under an SNR of 25 dB, the proposed algorithm can significantly improve the identification performance, achieving over 95% and approaching the ideal scenario of no receiver distortion. Meanwhile, influences caused by receiver distortions can be effectively eliminated, and the database can be shared with new receivers, overperforming other SEI methods that eliminate the receiver.

Mri findings, looking behaviour and affect recognition in very preterm children: A pilot study

Children born very preterm often exhibit atypical gaze behaviors, affect recognition difficulties and are at risk for cerebral white matter damage. This study explored links between these sequalae. In 24 12-year-old children born very preterm, ventricle size using Evans and posterior ventricle indices, and corpus callosum area were used to measure white matter thickness. The findings revealed a correlation between less attention towards the eyes and larger ventricle size. Ventricle and posterior corpus callosum sizes were correlated to affect-recognition proficiency. Findings suggest a link between white matter damage, gaze behavior, and affect recognition accuracy, emphasizing a relation with social perception.

Studies Advanced in License Plate Recognition

License plate recognition is a crucial mission in computer vision, and deep learning has significantly improved its performance. A representative license plate recognition system involves five components: license plate image preprocessing, image acquisition, license plate detection, character recognition, and character segmentation. This paper will explore the methods commonly used in each stage of the recognition process and analyze some of the current challenges and future trends of license plate recognition. This discussion will consider real-world factors such as lighting and weather conditions that can affect recognition accuracy. Ultimately, it is hoped that these insights will contribute to the development of intelligent transportation systems.

A dynamically optimized time-window length for SSVEP based hybrid BCI-VR system

Identification accuracy and information transfer rate are two critical performance indicators of the brain-computer interfaces and virtual reality (BCI_VR) hybrid system. A parameter closely related to these two indicators is the time-window length of the data. When the time-window value is too large, it is not conducive to the information transmission rate. Similarly, it affects recognition accuracy when the value is too small. How to set the time-window value reasonably to achieve a balance between recognition accuracy and information transfer rate to ensure the high-performance of the hybrid system is very important. In this paper, a dynamic time-window length optimized mechanism is proposed, which decides whether to change the time-window value of different subjects or trials according to the online performance of the hybrid system. Ten participants took part in the online experiment with an average accuracy of 92.07% and an average information transfer rate of 34.367 bits/min. Compared with the time-window adjustment method based on fixed-step in the existing study, the proposed method has achieved a 10.51% increase in information transfer rate and a 5.66% reduction in average detection time under the same experimental flow.

Multi-Window Identification of Landslide Hazards Based on InSAR Technology and Factors Predisposing to Disasters

Identification of potential landslide hazards is of great significance for disaster prevention and control. CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks) and many other deep learning methods have been used to identify landslide hazards. However, most samples are made with a fixed window size, which affects recognition accuracy to some extent. This paper presents a multi-window hidden danger identification CNN method according to the scale of the landslide in the experimental area. Firstly, the hidden danger area is preliminarily screened by InSAR deformation processing technology. Secondly, based on topography, geology, hydrology and human activities, a total of 15 disaster-prone factors are used to create factor datasets for in-depth learning. According to the general scale of the landslide, models with four window sizes of 48 × 48, 32 × 32, 16 × 16 and 8 × 8 are trained, respectively, and several window models with better recognition effect and suitable for the scale of landslide in the experimental area are selected for the accurate identification of landslide hazards. The results show that, among the four windows, 16 × 16 and 8 × 8 windows have the best model recognition effect. Then, according to the scale of the landslide, these optimal windows are pertinently selected, and the precision, recall rate and F-measure of the multi-window deep learning model are improved (82.86%, 78.75%, 80.75%). The research results prove that the multi-window identification method of landslide hazards combining InSAR technology and factors predisposing to disasters is effective, which can play an important role in regional disaster identification and enhance the scientific and technological support ability of geological disaster prevention and mitigation.

A Systematic Approach for Developing a Robust Artwork Recognition Framework Using Smartphone Cameras

The provision of information encourages people to visit cultural sites more often. Exploiting the great potential of using smartphone cameras and egocentric vision, we describe the development of a robust artwork recognition algorithm to assist users when visiting an art space. The algorithm recognizes artworks under any physical museum conditions, as well as camera point of views, making it suitable for different use scenarios towards an enhanced visiting experience. The algorithm was developed following a multiphase approach, including requirements gathering, experimentation in a virtual environment, development of the algorithm in real environment conditions, implementation of a demonstration smartphone app for artwork recognition and provision of assistive information, and its evaluation. During the algorithm development process, a convolutional neural network (CNN) model was trained for automatic artwork recognition using data collected in an art gallery, followed by extensive evaluations related to the parameters that may affect recognition accuracy, while the optimized algorithm was also evaluated through a dedicated app by a group of volunteers with promising results. The overall algorithm design and evaluation adopted for this work can also be applied in numerous applications, especially in cases where the algorithm performance under varying conditions and end-user satisfaction are critical factors.

Application of improved virtual sample and sparse representation in face recognition

AbstractSparse representation plays an important role in the research of face recognition. As a deformable sample classification task, face recognition is often used to test the performance of classification algorithms. In face recognition, differences in expression, angle, posture, and lighting conditions have become key factors that affect recognition accuracy. Essentially, there may be significant differences between different image samples of the same face, which makes image classification very difficult. Therefore, how to build a robust virtual image representation becomes a vital issue. To solve the above problems, this paper proposes a novel image classification algorithm. First, to better retain the global features and contour information of the original sample, the algorithm uses an improved non‐linear image representation method to highlight the low‐intensity and high‐intensity pixels of the original training sample, thus generating a virtual sample. Second, by the principle of sparse representation, the linear expression coefficients of the original sample and the virtual sample can be calculated, respectively. After obtaining these two types of coefficients, calculate the distances between the original sample and the test sample and the distance between the virtual sample and the test sample. These two distances are converted into distance scores. Finally, a simple and effective weight fusion scheme is adopted to fuse the classification scores of the original image and the virtual image. The fused score will determine the final classification result. The experimental results show that the proposed method outperforms other typical sparse representation classification methods.

Effects of Image Dataset Configuration on the Accuracy of Rice Disease Recognition Based on Convolution Neural Network.

In recent years, the convolution neural network has been the most widely used deep learning algorithm in the field of plant disease diagnosis and has performed well in classification. However, in practice, there are still some specific issues that have not been paid adequate attention to. For instance, the same pathogen may cause similar or different symptoms when infecting plant leaves, while the same pathogen may cause similar or disparate symptoms on different parts of the plant. Therefore, questions come up naturally: should the images showing different symptoms of the same disease be in one class or two separate classes in the image database? Also, how will the different classification methods affect the results of image recognition? In this study, taking rice leaf blast and neck blast caused by Magnaporthe oryzae, and rice sheath blight caused by Rhizoctonia solani as examples, three experiments were designed to explore how database configuration affects recognition accuracy in recognizing different symptoms of the same disease on the same plant part, similar symptoms of the same disease on different parts, and different symptoms on different parts. The results suggested that when the symptoms of the same disease were the same or similar, no matter whether they were on the same plant part or not, training combined classes of these images can get better performance than training them separately. When the difference between symptoms was obvious, the classification was relatively easy, and both separate training and combined training could achieve relatively high recognition accuracy. The results also, to a certain extent, indicated that the greater the number of images in the training data set, the higher the average classification accuracy.

Shade recognition of the color label based on the fuzzy clustering

In this article the task of determining the current position of pneumatic actuators is considered. The solution to the given task is achieved by using a technical vision system that allows to apply the fuzzy clustering method to determine in real time the center coordinates and the displacement position of a color label located on the mechatronic complex actuators. The objective of this work is to improve the accuracy of the moving actuator’s of mechatronic complex by improving the accuracy of the color label recognition. The intellectualization of process of the color shade recognition is based on fuzzy clustering. First, a fuzzy model is built, that allows depending on the input parameters of the color intensity for each of the RGB channels and the color tone component, to select a certain color in the image. After that, the color image is binarized and noise is suppressed. The authors used two defuzzification models during simulation a fuzzy system: one is based on the center of gravity method (CoG) and the other is based on the method of area ratio (MAR). The model is implemented based on the method of area ratio and allows to remove the dead zones that are present in the center of gravity model. The method of area ratio determines the location of the color label in the image frame. Subsequently, when the actuator is moved longitudinally, the vision system determines the location of the color label in the new frame. The color label position offset between the source and target images allows to determine the moved distance of the color label. In order to study how noise affects recognition accuracy, the following digital filters were used: median, Gaussian, matrix and binomial. Analysis of the accuracy of these filters showed that the best result was obtained when using a Gaussian filter. The estimation was based on the signal-to-noise coefficient. The mathematical models of fuzzy clustering of color label recognition were simulated in the Matlab/Simulink environment. Experimental studies of technical vision system performance with the proposed fuzzy clustering model were carried out on a pneumatic mechatronic complex that performs processing, moving and storing of details. During the experiments, a color label was placed on the cylinder, after which the cylinder moved along the guides in the longitudinal direction. During the movement, video recording and image recognition were performed. To determine the accuracy of color label recognition, the PSNR and RMSE coefficients were calculated which were equal 38.21 and 3.14, respectively. The accuracy of determining the displacement based on the developed model for recognizing color labels was equal 99.7%. The defuzzifier speed has increased to 590 ns.

Dynamic Segmentation for Physical Activity Recognition Using a Single Wearable Sensor

Data segmentation is an essential process in activity recognition when using machine learning techniques. Previous studies on physical activity recognition have mostly relied on the sliding window approach for segmentation. However, choosing a fixed window size for multiple activities with different durations may affect recognition accuracy, especially when the activities belong to the same category (i.e., dynamic or static). This paper presents and verifies a new method for dynamic segmentation of physical activities performed during the rehabilitation of individuals with spinal cord injuries. To adaptively segment the raw data, signal characteristics are analyzed to determine the suitable type of boundaries. Then, the algorithm identifies the time boundaries to represent the start- and endpoints of each activity. To verify the method and build a predictive model, an experiment was conducted in which data were collected using a single wrist-worn accelerometer sensor. The experimental results were compared with the sliding window approach, indicating that the proposed method outperformed the sliding window approach in terms of overall accuracy, which exceeded 5%, as well as model robustness. The results also demonstrated efficient physical activity segmentation using the proposed method, resulting in high classification performance for all activities considered.

Robust license plate detection and recognition with automatic rectification

We propose a robust license plate detection and recognition (LPDR) framework with automatic rectification. We explore the YOLOv2 object detector based on deep learning and train it to detect license plates (LPs) effectively. The LPs in natural scene images tend to be tilted and distorted because of the shooting angle or the geometric deformation of LPs. To solve the problem in which the LP tilt and distortion affect recognition accuracy, we introduce a spatial transformation network with thin-plate-spline transformation and propose a neural network called inverse compositional spatial transformer network-hierarchical spatial transformer network (ICSTN-CRNN). ICSTN-CRNN can automatically rectify and recognize LPs. Furthermore, we manually supplement the LP character annotations in PKUData. Our LPDR method achieves satisfactory results on three datasets, including Chinese City Parking Dataset, PKUData, and application-oriented license plate. Through a series of comparative experiments, we prove that our method is more accurate than other advanced methods.

Intentional Emotional Face Production in Children With Autism Spectrum Disorder Is Related to Social Functioning but Not Affect Recognition

Abstract Date Presented 03/28/20 The results of this study indicate the need to develop interventions targeting EFP abilities for ASD. Children with ASD showed decreased valence, but not accuracy, of their positive-valence facial expressions. In ASD, there was a negative correlation between EFP accuracy and ASD symptoms related to social functioning. There were no relationships found between affect recognition accuracy and EFP accuracy. Primary Author and Speaker: Sharada Krishnan Additional Authors and Speakers: Emily Kilroy Contributing Authors: Christiana Butera, Laura Harrison, Aditya Jayashankar, Anusha Hossain, Alexis Nalbach, Lisa Aziz-Zadeh

A comprehensive comparative study of handcrafted methods for face recognition LBP-like and non LBP operators

Pattern recognition and computer vision fields experienced the proposal of several architectures and approaches to deal with the demands of real world applications including face recognition. They have almost the same structure, based generally on a series of steps where the main ones are feature extraction and classification. The literature works interessted in face recognition problems, insist on the role of texture description as one of the key elements in face analysis, since it greatly affects recognition accuracy. Therefore, texture feature extraction has gained much attention and became a long-standing research topic thanks to its abilities to efficiently understand the face recognition process, especially in terms of face description. Recently, several literature researches in face application proposed new architectures based on pattern description proved by their discriminative power when extracting the feature information from facial images. These advantages combined with an outstanding performance in many classification applications, allowed the LBP-like descriptors to be one of the most prominent texture description method. Given this period of remarkable evolution, this research work includes a comprehensive analytical study of the face recognition performance of 64 LBP-like and 3 non-LBP texture descriptors recently proposed in the literature. To this end, we adopted a face recognition framework composed of four stages: 1) image pre-processing using gamma correction; 2) feature extraction using texture descriptors; 3) histogram calculation and 4) face recognition and classification based on the simple parameter-free Nearest Neighbors classifier (NN). The conducted comprehensive evaluations and experiments on the challenging and widely used benchmarks ORL, YALE, Extended YALE B and FERET databases presenting different challenges, indicate that a number of evaluated texture descriptors, which are tested for the first time on face recognition task, achieve better or competitive compared to several recent systems reported in face recognition literature.

A Survey on CSI-Based Human Behavior Recognition in Through-the-Wall Scenario

Recent years have witnessed increasing research interest in human behavior recognition as it provides attractive applications in various sensing scenarios. Among these encouraging implementations, device-free behavior recognition based on WiFi channel state information (CSI) has attracted significant attention due to the popularity of WiFi devices and abundant channel characteristics from CSI. Meanwhile, the CSI signal provides us with additional benefits because it can propagate through a wall. This through-the-wall and device-free scheme not only enables us to identify specific human actions but also to infer person activities by collecting the data from different rooms. This paper presents a survey on the state-of-art progresses in device-free through-the-wall human behavior recognition based on CSI. Specifically, this paper first introduces the basic concept of CSI and describes the signal variation caused by human behavior. Then, it illustrates that different human behaviors can cause signal transformation. Therefore, the unique map relationship between action and signal variation can be leveraged to recognize human behavior. Next, it provides the general architecture of through-the-wall behavior recognition and highlights its core characteristic. It investigates the state-of-art applications in various scenarios and analyzes specific design schemes and implementations. Afterward, it discusses the various across wall applications and makes a detailed comparison between non-through-the-wall and through-the-wall applications. Meanwhile, it analyzes many factors that affect recognition accuracy and emphasizes performance differences under across wall scenarios. Finally, this paper concludes by summarizing the issues and challenges faced and providing insights into the possible solution and future research trend.

The Association of Emotionally Congruent Responses with Emotion Recognition Accuracy after Brain Injury

To compare, in participants with a TBI and Healthy controls (HCs), emotional responses to affective film clips that were congruent with the emotions portrayed by the characters (i.e., felt the same emotion as the character) and examine associations of emotional congruence with affect recognition accuracy and empathy.

Double-Sided Occluded Chinese Character Recognition Accuracy and Response Time for Design and Nondesign Educational Background

This article investigates the influence of the position of occlusion, structural composition, and design educational status on Chinese character recognition accuracy and response time. Tsao and Liao conducted an experiment using 18 of the 4,000 most commonly used Chinese characters and suggested that the primary and secondary recognition features of a “single-sided” occluded Chinese character are the key radical (or initial strokes) and the key component (i.e., combination of strokes), respectively. The study concluded that right-side occluded characters require a shorter response time and yield more accurate recognition and that educational background does not significantly affect recognition accuracy and response time. The present study considered the same 18 Chinese characters and extended the work of Tsao and Liao by exploring accuracy rate and response time in design and nondesign educational groups for the recognition of “double-sided” occluded Chinese characters. The experimental results indicated that right-side occlusion (including both bottom-right and top-right occlusion) requires a shorter response time and yields more accurate recognition than left-side occlusion. These results agree with those of Tsao and Liao, who found that the key radical of a Chinese character is its key visual recognition feature. Even double-sided occlusion of Chinese characters does not affect the recognition outcome if the position of occlusion does not blur the key radical. Moreover, the participants majoring in design recognized the occluded Chinese characters more slowly than those with no educational background in design.

Sentence recognition in noise: The interplay between talker intelligibility, familiarity, and linguistic complexity

Foreign-accented speech recognition is typically tested with linguistically simple sentences, which offer a limited window into speech processing. In this study, participants transcribed simple (i.e. mono-clausal, canonical declarative syntax) and complex (i.e. multi-clausal, non-canonical syntax, and/or passive voice) sentences in noise at two signal-to-noise ratios. The sentences were spoken by three talkers: one native, one high and one low intelligibility L2 English speaker. Unsurprisingly, recognition accuracy dropped with decreasing intelligibility and signal-to-noise ratio. Interestingly, accuracy dropped with increasing linguistic complexity only for the native and high intelligibility talkers. For the low intelligibility talker, sentence complexity did not affect recognition accuracy. This pattern of findings was present in mixed and blocked experimental designs, indicating no effect of familiarity with the talker. We argue that listeners employ qualitatively different speech processing strategies for low versus high intelligibility talkers, and suggest that this difference may be driven by the talker’s realization of prosodic structure. While native and high intelligibility talkers effectively encode the contrast between complex and simple sentences, low intelligibility talkers may not convey sufficient prosodic information, resulting in the recognition of both sentence types as non-hierarchically organized word sequences. [Work supported by Knowles Hearing Center, NIH grants R01-DC012289 and R01-DC005794.]

Affect recognition from facial movements and body gestures by hierarchical deep spatio-temporal features and fusion strategy

Affect presentation is periodic and multi-modal, such as through facial movements, body gestures, and so on. Studies have shown that temporal selection and multi-modal combinations may benefit affect recognition. In this article, we therefore propose a spatio-temporal fusion model that extracts spatio-temporal hierarchical features based on select expressive components. In addition, a multi-modal hierarchical fusion strategy is presented. Our model learns the spatio-temporal hierarchical features from videos by a proposed deep network, which combines a convolutional neural networks (CNN), bilateral long short-term memory recurrent neural networks (BLSTM-RNN) with principal component analysis (PCA). Our approach handles each video as a “video sentence.” It first obtains a skeleton with the temporal selection process and then segments key words with a certain sliding window. Finally, it obtains the features with a deep network comprised of a video-skeleton and video-words. Our model combines the feature level and decision level fusion for fusing the multi-modal information. Experimental results showed that our model improved the multi-modal affect recognition accuracy rate from 95.13% in existing literature to 99.57% on a face and body (FABO) database, our results have been increased by 4.44%, and it obtained a macro average accuracy (MAA) up to 99.71%.

Automatic identification of fungi in microscopic leucorrhea images.

Identifying fungi in microscopic leucorrhea images provides important information for evaluating gynecological diseases. Subjective judgment and fatigue can greatly affect recognition accuracy. This paper proposes an automatic identification system to detect fungi in leucorrhea images that incorporates a convolutional neural network, the histogram of oriented gradients algorithm, and a binary support vector machine. In experiments, the detection rate of the positive samples was as high as 99.8%. The experimental results demonstrate the effectiveness of the proposed method and its potential as a primary software component of a completely automated system.