This paper presents a novel feature point descriptor for the multispectral image case: Far-Infrared and Visible Spectrum images. It allows matching interest points on images of the same scene but acquired in different spectral bands. Initially, points of interest are detected on both images through a SIFT-like based scale space representation. Then, these points are characterized using an Edge Oriented Histogram (EOH) descriptor. Finally, points of interest from multispectral images are matched by finding nearest couples using the information from the descriptor. The provided experimental results and comparisons with similar methods show both the validity of the proposed approach as well as the improvements it offers with respect to the current state-of-the-art.

This paper presents a novel CNN-based architecture, referred to as Q-Net, to learn local feature descriptors that are useful for matching image patches from two different spectral bands. Given correctly matched and non-matching cross-spectral image pairs, a quadruplet network is trained to map input image patches to a common Euclidean space, regardless of the input spectral band. Our approach is inspired by the recent success of triplet networks in the visible spectrum, but adapted for cross-spectral scenarios, where, for each matching pair, there are always two possible non-matching patches: one for each spectrum. Experimental evaluations on a public cross-spectral VIS-NIR dataset shows that the proposed approach improves the state-of-the-art. Moreover, the proposed technique can also be used in mono-spectral settings, obtaining a similar performance to triplet network descriptors, but requiring less training data.

We report a novel proposal for reducing the digital divide in rural multigrade schools, incorporating knowledge of robotics with a STEM approach to simultaneously promote curricular learning in mathematics and science in several school grades. We used an exploratory qualitative methodology to implement the proposal with 12 multigrade rural students. We explored the contribution of the approaches to the promotion of curricular learning in mathematics and science and the perceptions of using robotics to learn mathematics and science. As data collection techniques, we conducted focus groups and semi-structured interviews with the participants and analyzed their responses thematically. We concluded that the proposal could contribute to meeting the challenges of multigrade teaching. Our findings suggest that the proposal would simultaneously promote the development of curricular learning in mathematics and science in several school grades, offering an alternative for addressing various topics with different degrees of depth.

In blueberry farming, accurately assessing maturity is critical to efficient harvesting. Deep Learning solutions, which are increasingly popular in this area, often undergo evaluation through metrics like mean average precision (mAP). However, these metrics may only partially capture the actual performance of the models, especially in settings with limited resources like those in agricultural drones or robots. To address this, our study evaluates Deep Learning models, such as YOLOv7, RT-DETR, and Mask-RCNN, for detecting and classifying blueberries. We perform these evaluations on both powerful computers and embedded systems. Using Type-Influence Detector Error (TIDE) analysis, we closely examine the accuracy of these models. Our research reveals that partial occlusions commonly cause errors, and optimizing these models for embedded devices can increase their speed without losing precision. This work improves the understanding of object detection models for blueberry detection and maturity estimation.

This article introduces a voice-controlled robotic system for early education, enabling children as young as four to interact with robots using natural voice commands. Recognizing the challenges posed by programming languages and robot theory for young learners, this study leverages recent advancements in artificial intelligence, such as large language models, to make robots more intelligent and easier to use. This innovative approach fosters a natural and intuitive interaction between the child and the robot, effectively removing barriers to access and expanding the educational possibilities of robotics in the classroom. In this context, a software pipeline is proposed that translates voice commands into robot actions. Each component is tested using different deep learning models and cloud services to determine their suitability, with the best ones being selected. Finally, the chosen setup is validated through an integration test involving children aged 4 to 6 years. Preliminary results demonstrate the system’s capability to accurately recognize and execute voice commands, highlighting its potential as a valuable educational tool for early education.

Current CNN-based stereo depth estimation models can barely run under real-time constraints on embedded graphic processing unit (GPU) devices. Moreover, state-of-the-art evaluations usually do not consider model optimization techniques, being that it is unknown what is the current potential on embedded GPU devices. In this work, we evaluate two state-of-the-art models on three different embedded GPU devices, with and without optimization methods, presenting performance results that illustrate the actual capabilities of embedded GPU devices for stereo depth estimation. More importantly, based on our evaluation, we propose the use of a U-Net like architecture for postprocessing the cost-volume, instead of a typical sequence of 3D convolutions, drastically augmenting the runtime speed of current models. In our experiments, we achieve real-time inference speed, in the range of 5–32 ms, for 1216 × 368 input stereo images on the Jetson TX2, Jetson Xavier, and Jetson Nano embedded devices.