It is difficult for optical correlator to recognize the target with large rotation distortion by using existing filters. To solve this problem, a new neural network model is constructed based on the physical recognition process of planar integrated 2F optical correlator. The new optical filter is optimized by training the neural network. The planar integrated 2f optical correlator can output sharp correlation peaks by using the new filter to recognize targets with arbitrary rotation distortion. Compared with the traditional OTSDF filter, the average increase of correlation peak index is up to 1402.4%, while the recognition performance of distortion invariant is also better than the new NNCRF filter. The simulation and experimental results show that the optical filter designed in this paper can effectively solve the problem of weak recognition ability of optical correlator for rotating distortion targets, especially for large angle rotating distortion targets.

Natural disasters create complex environments where effective human detection is both critical and challenging, especially when individuals are partially occluded. While recent advancements in computer vision have improved detection capabilities, there remains a significant need for efficient solutions that can enhance search-and-rescue (SAR) operations in resource-constrained disaster scenarios. This study modified the original DINO (Detection Transformer with Improved Denoising Anchor Boxes) model and introduced the visibility-enhanced DINO (VE-DINO) model, designed for robust human detection in occlusion-heavy environments, with potential integration into SAR system. VE-DINO enhances detection accuracy by incorporating body part key point information and employing a specialized loss function. The model was trained and validated using the COCO2017 dataset, with additional external testing conducted on the Disaster Occlusion Detection Dataset (DODD), which we developed by meticulously compiling relevant images from existing public datasets to represent occlusion scenarios in disaster contexts. The VE-DINO achieved an average precision of 0.615 at IoU 0.50:0.90 on all bounding boxes, outperforming the original DINO model (0.491) in the testing set. The external testing of VE-DINO achieved an average precision of 0.500. An ablation study was conducted and demonstrated the robustness of the model subject when confronted with varying degrees of body occlusion. Furthermore, to illustrate the practicality, we conducted a case study demonstrating the usability of the model when integrated into an unmanned aerial vehicle (UAV)-based SAR system, showcasing its potential in real-world scenarios.

Accurate, unobtrusive in-bed occupancy-posture Internet of Thing (IoT) monitoring is critical both for timely detection of “missing patient” events, reducing false alarms in nocturnal surveillance, and informing sleep posture-related health risk management. This article presents a novel two-stage, multitask, multiple ultrawideband (UWB) radar IoT system for joint in-bed occupancy detection and fine-grained sleep posture classification, using an array of eight UWB radars. The technical novelty centers on three mechanisms: a view-weighted multiradar network (VWM-Net) that adaptively fuses single-radar echo maps into a spatial echo feature map (SEFM); an integration of DenseNet and ConvNeXt2 (DCNX2); and a cross-modal generative model that translates SEFMs into human-interpretable depth posture images. Specifically, a UCYC-GAN is proposed by integrating a U-Net structure with a CycleGAN generator, and its visual utility is compared against NICE-GAN and bidirectional diffusion bridge model (BDBM). The system is trained and validated on a cohort of 100 participants across ten fine-grained postures and four blanket conditions, supplemented with a life-sized dummy and cluttered objects to model nonhuman occupancy and environmental distractors. In the first stage, a dedicated classifier distinguishes humans from nonhuman entities with 99.09% accuracy. In the second stage, the multihead DCNX2 backbone performed multitask classification, achieving 93.75% accuracy for ten fine-grained postures, 98% for four coarse-grained postures, 88.25% for blanket-coverage types, 98.13% for blanket presence, and 95.13% for participant gender. The overall can effectively mitigate false positives from environmental clutter, advances noncontact in-bed occupancy-posture surveillance, and, through a cloud-enabled web dashboard and APIs, supports real-time visualization and retrospective data analysis.

This chapter introduces various machine learning methods used for predicting functional recovery and prognosis in conditions such as Parkinson’s disease and dementia. It covers a range of models, starting with traditional approaches like support vector machines (SVMs), gradient boosting machines (GBMs), and its variant extreme gradient boosting (XGB), which are useful for classification. Simpler methods such as naïve Bayes classifier (NBC) and K-nearest Neighbors (KNNs) are explored for their nonparametric characteristics, while multiple linear regression (MLR) is discussed for modeling linear relationships. Deep learning (DL) models like convolutional neural networks (CNNs) and artificial neural networks (ANNs) are also introduced for capturing complex data patterns. Lasso regression (LR) is highlighted for its utility in feature selection. The chapter discusses the limitations of these models, such as interpretability and computational demands, and considers potential improvements through integrating multiple models to enhance clinical prediction accuracy. <!----> <!----> <!----> <!---->