Learning interest has been found to have insignificant and significant relationships with academic performance among students. This study aims to examine the association between learning interest and academic achievement among undergraduate students who underwent a Hybrid-flexible (HyFlex) course which was designed and implemented based on the interest loop of Interest-driven Creator (IDC) theory. In the HyFlex learning setting, students had the autonomy to learn physically, online synchronously, or online asynchronously for each class session. 48 undergraduate students from the faculty of education at a public university in Malaysia participated in this study. Pearson correlation analysis shows a significant positive correlation between learning interest and academic performance. A simple linear regression analysis found that learning interest contributed significantly to academic performance. This implies that the integration of the IDC theory into HyFlex learning could have the potential to foster interest and enhance performance among undergraduate students. However, the small effect size of learning interest indicates that while it is pertinent to cultivate learning interest among learners, academic achievement could be affected by other factors beyond interest.

In this paper, the authors adopted a mixed methods explanatory sequential approach to examine student teachers’ satisfaction and its relationship with participation modalities in a HyFlex course. It also explored the factors behind their extent of satisfaction and choice of participation modalities. Twenty-six student teachers completed the satisfaction questionnaire, and four participants were recruited for a focus group discussion. The research findings indicated that students were very satisfied with the HyFlex course because the flexibility in attendance and the easy access to learning materials accommodated their learning needs and personal commitments; this promoted self-directed learning and reduced their academic pressure. The student teachers were also satisfied with the HyFlex course because they gained exposure to an innovative learning format. As for the factors affecting student choice of participation modalities in the HyFlex course, peer influence is one of the significant factors. Students also reported that they chose the synchronous mode based on their learning preferences and empathy towards the instructor's commitment, while they chose the online options due to their energy level, health conditions, and personal matters. Overall, this study suggests that HyFlex may be considered as a learning model that could cater to various learning needs among student teachers majoring in baccalaureate in education (physical education), and potentially enhance their learning outcomes.

This paper discusses the concept of seamless learning and its alignment with Interest-Driven Creator (IDC) Theory in response to Tak-Wai Chan’s keynote at the 1st MetaACES 2023. He introduced the Seamless Interest-Driven Co-Creator (SIDC) learning framework — an integration of seamless learning and IDC learning. This paper also explores the educational implications of SIDC learning framework.

Traditional monocular visual simultaneous localization and mapping (SLAM) algorithms have been extensively studied and proven to reliably recover a sparse structure and camera motion. Nevertheless, the sparse structure is still insufficient for scene interaction, e.g., visual navigation and augmented reality applications. To densify the scene reconstruction, the use of single-image absolute depth prediction from convolutional neural networks (CNNs) for filling in the missing structure has been proposed. However, the prediction accuracy tends to not generalize well on scenes that are different from the training datasets.In this paper, we propose a dense monocular SLAM system, named DeepRelativeFusion, that is capable to recover a globally consistent 3D structure. To this end, we use a visual SLAM algorithm to reliably recover the camera poses and semi-dense depth maps of the keyframes, and then use relative depth prediction to densify the semi-dense depth maps and refine the keyframe pose-graph. To improve the semi-dense depth maps, we propose an adaptive filtering scheme, which is a structure- preserving weighted average smoothing filter that takes into account the pixel intensity and depth of the neighbouring pixels, yielding substantial reconstruction accuracy gain in densification. To perform densification, we introduce two incremental improvements upon the energy minimization framework proposed by DeepFusion: (1) an improved cost function, and(2) the use of single-image relative depth prediction. After densification, we update the keyframes with two-view consistent optimized semi-dense and dense depth maps to improve pose- graph optimization, providing a feedback loop to refine the keyframe poses for accurate scene reconstruction. Our system outperforms the state-of-the-art dense SLAM systems quantitatively in dense reconstruction accuracy by a large margin.For more information, see the demo video and supplementary material.

Reliable feature correspondence between frames is a critical step in visual odometry (VO) and visual simultaneous localization and mapping (V-SLAM) algorithms. In comparison with existing VO and V-SLAM algorithms, semi-direct visual odometry (SVO) has two main advantages that lead to state-of-the-art frame rate camera motion estimation: direct pixel correspondence and efficient implementation of probabilistic mapping method. This paper improves the SVO mapping by initializing the mean and the variance of the depth at a feature location according to the depth prediction from a single-image depth prediction network. By significantly reducing the depth uncertainty of the initialized map point (i.e., small variance centred about the depth prediction), the benefits are twofold: reliable feature correspondence between views and fast convergence to the true depth in order to create new map points. We evaluate our method with two outdoor datasets: KITTI dataset and Oxford Robotcar dataset. The experimental results indicate that improved SVO mapping results in increased robustness and camera tracking accuracy. The implementation of this work is available at https: //github.com/yan99033/CNN-SVO

Curve-fitting problems are widely solved using numerical and soft techniques. In particular, artificial neural networks (ANN) are used to approximate arbitrary input–output relationships in the form of tuned edge weights. Moreover, using semantic networks such as fuzzy cognitive map (FCM), single graph nodes could be directly associated with their actual grey scales rather than binary values as in ANN. This article examines a novel methodology for automatic construction of FCMs for function approximation. The main contribution is the introduction of nested-FCM structure for multi-variable curve fitting. There are step-by-step example cases along with the obtained results to serve as a guide to the new methods being introduced. It is shown that nested FCM derives relationship models of multiple variables using any conventional weight training technique with minimal computation effort. Issues about computational cost and accuracy are also discussed along with future direction of the research.

Artificial potential field (APF) is well established for reactive navigation of robots. Initially, this paper describes a fast and robust fuzzy-APF on an ActivMedia AmigoBot platform. Obstacle-related information is fuzzified just by sensory fusion which results in shorter runtime. The membership functions of obstacles' range and direction have been also merged into one function for smaller block of rules. The fuzzy-APF is tested and verified virtually with non-concave obstacles. Main contribution of this article is a new approach to motion modeling. The goal is to discover decision making behaviors of the robot in wayfinding. A novel decision modeling technique is developed based on capabilities of the fuzzy cognitive map (FCM) and supervised learning using the genetic algorithm (GA). Decision productions for moving from one sub-space to another are modeled in form of decision matrices. The robot trajectory under supply of such decision matrices has likelihood of nearly 90% with its trajectory under the APF. Replication of robot motion is therefore achieved by modeling its decision behaviors in form of tangible matrices.

Today, collaborative learning in high-tech institutions is not limited to supply of material by instructors and merely its consumption by students. Instead, learning is viewed as a cooperative task involving group discussions and analysis of the supplied information. Group discussion takes different forms starting from primary school activities till collaborative research and development projects in universities. On the basis of this idea, this paper introduces a practical computerized model for analyzing different viewpoints of a group of individuals on a specific decision problem or a single fact of knowledge. The results are presented using an example of a decision problem including 10 factors where 3 decision makers discuss to reach a consensus over the finalized values of the factors.