Hyperspectral Face Recognition with Adaptive and Parallel SVMs in Partially Hidden Face Scenarios.

Hyperspectral imaging opens up new opportunities for masked face recognition via discrimination of the spectral information obtained by hyperspectral sensors. In this work, we present a novel algorithm to extract facial spectral-features from different regions of interests by performing computer vision techniques over the hyperspectral images, particularly Histogram of Oriented Gradients. We have applied this algorithm over the UWA-HSFD dataset to extract the facial spectral-features and then a set of parallel Support Vector Machines with custom kernels, based on the cosine similarity and Euclidean distance, have been trained on fly to classify unknown subjects/faces according to the distance of the visible facial spectral-features, i.e., the regions that are not concealed by a face mask or scarf. The results draw up an optimal trade-off between recognition accuracy and compression ratio in accordance with the facial regions that are not occluded.

Analytical Model for the Duty Cycle in Solar-Based EH-WSN for Environmental Monitoring.

A technology drift is currently taking place from traditional battery-powered sensor networks, which exhibit limited lifetime, to the new Energy-Harvesting Wireless Sensor Networks (EH-WSN), which open the way towards self-sustained operation. However, this emergent modality also brings up new challenges, especially due to the time-varying nature and unpredictability of ambient energy sources. Most proposals for implementing EH-WSN rely on heuristic approaches to redesign the duty-cycling mechanism at the MAC layer, with the ultimate goal of optimizing network performance while preserving self-sustained and continuous operation. In contrast to the common system-wide reduced duty cycle of battery-powered sensor networks, the duty cycle in EH-WSN is much larger and adapted to the energy harvesting rate and traffic load of each node in the network. In this paper, we focus on solar-based EH-WSN devoted to environmental monitoring. In contrast to current works, we follow an analytical approach, which results into closed-form expressions for the duty cycle and initial energy storage that guarantee self-sustained operation to any node in a solar-based EH-WSN. To center the analysis, we consider TinyOS sensor nodes, though we postulate that the essential components of the obtained formulation will contribute to further develop duty cycle adaptation schemes for TinyOS and other software platforms.

Cross Layer Adaptation of Check intervals in low power listening MAC protocols for lifetime improvement in Wireless Sensor Networks.

Preamble sampling-based MAC protocols designed for Wireless Sensor Networks (WSN) are aimed at prolonging the lifetime of the nodes by scheduling their times of activity. This scheduling exploits node synchronization to find the right trade-off between energy consumption and delay. In this paper we consider the problem of node synchronization in preamble sampling protocols. We propose Cross Layer Adaptation of Check intervals (CLAC), a novel protocol intended to reduce the energy consumption of the nodes without significantly increasing the delay. Our protocol modifies the scheduling of the nodes based on estimating the delay experienced by a packet that travels along a multi-hop path. CLAC uses routing and MAC layer information to compute a delay that matches the packet arrival time. We have implemented CLAC on top of well-known routing and MAC protocols for WSN, and we have evaluated our implementation using the Avrora simulator. The simulation results confirm that CLAC improves the network lifetime at no additional packet loss and without affecting the end-to-end delay.