Neural Differential Equations (NDEs) excel at modeling continuous-time dynamics, effectively handling challenges such as irregular observations, missing values, and noise. Despite their advantages, NDEs face a fundamental challenge in adopting dropout, a cornerstone of deep learning regularization, making them susceptible to overfitting. To address this research gap, we introduce Continuum Dropout, a universally applicable regularization technique for NDEs built upon the theory of alternating renewal processes. Continuum Dropout formulates the on-off mechanism of dropout as a stochastic process that alternates between active (evolution) and inactive (paused) states in continuous time. This provides a principled approach to prevent overfitting and enhance the generalization capabilities of NDEs. Moreover, Continuum Dropout offers a structured framework to quantify predictive uncertainty via Monte Carlo sampling at test time. Through extensive experiments, we demonstrate that Continuum Dropout outperforms existing regularization methods for NDEs, achieving superior performance on various time series and image classification tasks. It also yields better-calibrated and more trustworthy probability estimates, highlighting its effectiveness for uncertainty-aware modeling.

Time series modeling and analysis have become critical in various domains. Conventional methods such as RNNs and Transformers, while effective for discrete-time and regularly sampled data, face significant challenges in capturing the continuous dynamics and irregular sampling patterns inherent in real-world scenarios. Neural Differential Equations (NDEs) represent a paradigm shift by combining the flexibility of neural networks with the mathematical rigor of differential equations. This paper presents a comprehensive review of NDE-based methods for time series analysis, including neural ordinary differential equations, neural controlled differential equations, and neural stochastic differential equations. We provide a detailed discussion of their mathematical formulations, numerical methods, and applications, highlighting their ability to model continuous-time dynamics. Furthermore, we address key challenges and future research directions. This survey serves as a foundation for researchers and practitioners seeking to leverage NDEs for advanced time series analysis.

This paper proposes a MIMO imaging method with an iterative-based super-resolution technique for automotive radar applications. Vehicle radars have recently used 4D imaging radar, offering improved detection ranges and high-resolution capability. The application of imaging radar technology aims to extend the maximum detection distance through noise reduction techniques while also enabling the miniaturization of vehicle radars using the MIMO approach. To enhance the maximum detection distance, we employ a Wavelet-based noise reduction method for range FFT. Additionally, for improved angular resolution, we use a MIMO radar implementation based on a super-resolution algorithm, in contrast to the conventional MIMO imaging method that utilizes the FFT algorithm. Specifically, we focus on an emerging iterative-based algorithm, which effectively addresses the complexity issues associated with super-resolution techniques. Through extensive experiments, we validate the effectiveness of this proposed method. The results demonstrate its potential in realizing wide detection distance and high-resolution for vehicle radar systems.

This paper presents development of 24GHz millimeter wave smart radar for intelligent street lighting system. The developed radar operates at a frequency of 24GHz with 200MHz bandwidth and CW (continuous wave) mode. The two radars are used to cover the street in both directions and detect obstacles moving at the speed of more than 1 km/h including moving pedestrians. The radar detection controls the street lighting. Therefore, the energy-saving performance has been improved because the proposed smart lighting system works only when obstacles around the streetlight exist.

Object detection has been researched using a camera, a LIDAR and a RADAR. However, camera-based techniques have heavy image processing and are sensitive for light intensity. LIDAR can measure precise distance from objects, but it is difficult to classify objects. Further, it is known that when pedestrians are partially occluded, detecting them is extremely difficult because of insufficient data to determine them. To address this problem, we use LIDAR and RADAR sensors to improve the detection accuracy. We present a sensor fusion scheme for detecting partially occluded pedestrian with low-complexity.

Wireless video sensor networks (WVSNs) require video data from sensor nodes to be delivered efficiently. Cameras in adjacent video nodes tend to have correlated fields of view (FoVs) or overlapping part when a sufficient number of video sensors are deployed. This paper proposes a data aggregation technique for WVSNs to remove the spatial redundancy, thereby reducing energy consumption and response time. Our approach exploits the correlation between discrete cosine transform (DCT) coefficients pairs of two intra-coded images from cameras with overlapping FoVs. Experiments show that an intermediate node en route to the base station achieves bit-rate savings up to 18.9%. This scheme is less complicated than other video and image coding techniques that exploit correlated FoV, allowing resource-constrained video sensors to operate more reliably and longer.