Pedestrian detection is essential to avoid dangerous traffic situations. We present a fast and robust algorithm for detecting pedestrians in cluttered scene from pair of moving cameras. This achieved through stereo-based segmentation neural network-based recognition. The includes three steps. First, we segment the image into sub-image object candidates using disparities discontinuity. Second, merge split sub-images that satisfy pedestrian size shape constraints. Third, use intensity...

Textual data, as a key carrier of learning feedback, is continuously produced by many students within course forums. The temporal nature discussion requires students' emotions and concerned aspects (e.g. teaching styles, activities, etc.) to be dynamically tracked for understanding requirements. To characterize dynamics emotion-aspects, this paper presents an unsupervised model, namely emotion-aspect model (TEAM), modeling time jointly with capture evolutions over time. Especially, the...

Digital data trails from disparate sources covering different aspects of student life are stored daily in most modern university campuses. However, it remains challenging to (i) combine these obtain a holistic view student, (ii) use accurately predict academic performance, and (iii) such predictions promote positive engagement with the university. To initially alleviate this problem, article, model named Augmented Education (AugmentED) is proposed. In our study, (1) first, an experiment...

Known as the third revolution of information technology, Internet Things (IoT) embodies transformation human technology from "virtual" to "reality." The application IoT in education has risen concerns both researchers and practitioners. However, there are few research on using bibliometric visually analyze hotspots trends education. In this study, a total 2257 articles, including (i) 1243 domestic articles 2005 2021; (ii) 1014 foreign 2021, were collected for comparative analysis between...

As a sub-challenge of EmotiW (the Emotion Recognition in the Wild challenge), how to improve performance on AFEW (Acted Facial Expressions wild) dataset is popular benchmark for emotion recognition tasks with various constraints, including uneven illumination, head deflection, and facial posture. In this paper, we propose convenient expression cascade network comprising spatial feature extraction, hybrid attention, temporal extraction. First, video sequence, faces each frame are detected,...

Recent teaching trends are increasingly integrating diverse multimedia and computer-aided practices to enhance representation understanding. Leveraging data-focused strategies, these methods further refined with artificial intelligence decision-making techniques. To improve data handling in multimedia-based representation, this article introduces an integrated model aided by regression learning (IDR-RL). The proposed satisfies the required for different methods/ subjects based on curriculum...

Spiking Neural Networks (SNNs), which offer exceptional energy efficiency for inference, and Federated Learning (FL), offers privacy-preserving distributed training, is a rising area of interest that highly beneficial towards Internet Things (IoT) devices. Despite this, research tackles Byzantine attacks bandwidth limitation in FL-SNNs, both poses significant threats on model convergence training times, still remains largely unexplored. Going beyond proposing solution these problems, this...

Academic stress is a common psychological issue among college students, but the mechanism by which it affects academic performance not well understood. This study extends existing literature exploring chain-mediated model composed of stress, learning motivation, depression, and performance. The involved 591 students from university in Hubei Province, China, employed structural equation modelling to analyse data. results showed that did have significant direct effect on Instead, depression...

The application of information and communications technology (ICT) in higher educational institutions has led to the transformation environment from digital smart. assessment smart learning environments will highlight advantages disadvantages its construction results help establish a sustainable space for students' personalized study. However, level effect are difficult judge thoroughly, so comprehensive evaluation method is needed. By rethinking structure (the physical space, resource...

As fruit firmness is a crucial characteristic associated with the maturity level, its accurate estimation of great importance to post-harvest processing and wholesale in industry. Benefiting from advances soft robotics, gripper simultaneous compliant deformation tactile sensing proposed this study for classification. The design inspired by fin ray effect can achieve active deformation, which helps simplify actuation system improve delicate manipulation capability. Finite element modelling,...

We propose a closely coupled object detection and segmentation algorithm for enhancing both processes in cooperative iterative manner. Figure-ground reduces the effect of background clutter on template matching; matched provides shape constraints segmentation. More precisely, we estimate probability each pixel belonging to foreground by weighted sum estimates based color alone. The weight shape-based is related that familiar present updated dynamically so enforce only where present....

The performance of a facial expression recognition network degrades obviously under situations uneven illumination or partial occluded face as it is quite difficult to pinpoint the attention hotspots on dynamically changing regions (e.g., eyes, nose, and mouth) precisely possible. To address above issue, by hybrid mechanism pyramid feature, this paper proposes cascade attention-based basis combination (i) local spatial (ii) multi-scale-stereoscopic context feature (extracted from 3-scale...

Understanding learners' emotions can help optimize instruction sand further conduct effective learning interventions. Most existing studies on student emotion recognition are based multiple manifestations of external behavior, which do not fully use physiological signals. In this context, the one hand, a EEG dataset (LE-EEG) is constructed, captures signals reflecting boredom, neutrality, and engagement during learning; other an classification network attention fusion (ECN-AF) proposed. To...

Stress has significant effects on an individual's daily life in modern society, making its detection a topic of great interest over the decade. While numerous studies have delved into this field, accuracy and reliability stress methods still room for improvement. In study, we propose multimodal multitemporal-scale fusion-based system. First, hybrid feature extraction module is proposed, which generates set from perspective handcrafted deep learning (DL) analysis across multiple temporal...

We describe a system for human body pose estimation from multiple views that is fast and completely automatic. The algorithm works in the presence of people by decoupling problems different people. estimated based on likelihood function integrates information thus obtains globally optimal solution. Other characteristics make our method more general than previous work include: (1) no manual initialization; (2) specification dimensions 3D structure; (3) reliance some learned poses or patterns...

Certificate validation in Secure Sockets Layer or Transport Security protocol (SSL/TLS) is critical to Internet security. Thus, it significant check whether certificate SSL/TLS implementations correctly implemented. With this motivation, we propose a novel differential testing approach that based on the standard Request for Comments (RFC). First, rules of certificates are extracted automatically from RFCs. Second, low-level test cases generated through dynamic symbolic execution. Third,...

Drowsiness can lead to inefficiency and major disasters; thus it is important address in both academia industry. Despite multiple types of research this field, a nonintrusive classifier system for detecting drowsiness real-time under natural environment without specific stimulation lacking. This study develops detection using 77 GHz millimeter-wave (mmWave) frequency-modulated continuous wave radar. Specifically, firstly, mmWave processing module proposed, which adaptively suppress...

Estimation of human head orientation is important for a number applications such as human-computer interaction, teleconferencing, virtual reality, and 3D audio rendering. We present system estimating based on visual information. Two neural networks are trained to approximate the functions that map an image head. obtain ground-truth data training testing from electromagnetic tracking device worn by subjects. Our experimental results demonstrate accuracy within 10/spl deg/ with subject free...