Student response systems (SRSs) have generated significant debate and discussion in the educational research literature in the past decade. It is well known that they offer several important advantages including encouraging student interaction, offering anonymous and instant student feedback and improving the student learning experience. While several different types of such systems exist, they nevertheless remain limited in their input capabilities. In particular, they typically only allow for multiple-choice style responses, although some devices do cater for numerical and textual input. However, most of the available SRSs do not allow for freeform input such as mathematical equations, graphical drawings or circuit diagrams. This is of particular relevance to Engineering and Science disciplines where such information is core to the student learning. The approach to solving a problem is often as important, if not more so, than the actual final answer itself. This paper presents a classroom response system that allows for freeform input and operates on any smart media device with a touch interface and that employs the Android operating system, such as everyday smart phones and tablets. The proposed system involves three different components, namely a student application that allows for sketch capabilities, a lecturer application that allows for the viewing and marking of multiple student sketches and a cloud service for the exchange of messages. In addition, this proposed system was evaluated by a class of engineering students at NUI Maynooth, the results of which are presented within.
This paper presents a benchmark data set for evaluating ball detection algorithms in the RoboCup Soccer Standard Platform League. We created a labelled data set of images with and without ball derived from vision log files recorded by multiple NAO robots in various lighting conditions. The data set contains 5209 labelled ball image regions and 10924 non-ball regions. Non-ball image regions all contain features that had been classified as a potential ball candidate by an existing ball detector. The data set was used to train and evaluate 252 different Deep Convolutional Neural Network (CNN) architectures for ball detection. In order to control computational requirements, this evaluation focused on networks with 2-5 layers that could feasibly run in the vision and cognition cycle of a NAO robot using two cameras at full frame rate (2×30 Hz). The results show that the classification performance of the networks is quite insensitive to the details of the network design including input image size, number of layers and number of outputs at each layer. In an effort to reduce the computational requirements of CNNs we evaluated XNOR-Net architectures which quantize the weights and activations of a neural network to binary values. We examined XNOR-Nets corresponding to the real-valued CNNs we had already tested in order to quantify the effect on classification metrics. The results indicate that ball classification performance degrades by 12% on average when changing from real-valued CNN to corresponding XNOR-Net.
Classroom assessment techniques (CATs) are ungraded activities in a classroom setting that provide feedback to the teacher and to the students themselves, on the current state of student learning and understanding, which can subsequently drive corrective actions where necessary. Student response systems (SRSs) provide a technological solution for CATs whereby students can respond anonymously and instructors can provide instant feedback. However, existing systems have tended to suffer from constrained input, limiting the quality of the student responses. In particular, existing SRSs typically only employ well known form based input metaphors such as the multiple-choice selection and text-box input. These input types are not well suited to responses that require significant graphic or symbolic elements such as equations, circuit diagrams, and other drawings. These SRSs also have logistical issues in relation to portability and ownership of the equipment. In this paper, the authors present an SRS designed from the ground up to support CATs with freeform input to fulfil the needs of the science, technology, engineering and mathematics (STEM) classroom, though the solution is applicable to any learning environment in which freeform input is valuable. To mitigate logistical issues, the solution employs touch based Android tablets and smart phones commonly owned by students and a freely downloadable student app. This paper details the design of teacher and student interaction, including instructor preparation prior to class. The authors also examine some of the issues surrounding freeform graphic and symbolic input on a range of device form factors and the particular solutions that they found effective. A summary of their ongoing evaluation of this system is also outlined within.
Abstract-In multi-robot control, robots need the ability to perform well in the absence of valid human input. This paper describes a shared control scheme for multiple robots where control can be traded between human and autonomous agents on the fly, reducing the negative effect of robots requiring attention while the user's attention is devoted elsewhere. The control approach is a hybrid of traded control where the control is traded between the human and autonomous agents and coordinated control where high level human user commands are mapped to low level implementation under partial control of the autonomous system. An allocation authority decides whether the user or autonomous agent is controlling the robot at a given time based on a command context. This allocation authority control scheme is compared to a coordinated control scheme in a multirobot soccer domain. The results demonstrate that an allocation authority approach produces improved task performance and may have generalizable applicability. Specifically in the context of robotic soccer, the proposed control scheme was shown to have 18.5% more possession and 8.2% more territory than traded control while also reducing mental demands on users.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.