Fast-moving Objects Research Articles

Neural-Rendezvous: Provably Robust Guidance and Control to Encounter Interstellar Objects

Interstellar objects (ISOs) are likely representatives of primitive materials invaluable in understanding exoplanetary star systems. Due to their poorly constrained orbits with generally high inclinations and relative velocities, however, exploring ISOs with conventional human-in-the-loop approaches is significantly challenging. This paper presents Neural-Rendezvous—a deep-learning-based guidance and control framework for encountering fast-moving objects, including ISOs, robustly, accurately, and autonomously in real time. It uses pointwise minimum norm tracking control on top of a guidance policy modeled by a spectrally normalized deep neural network, where its hyperparameters are tuned with a loss function directly penalizing the model predictive control state trajectory tracking error. We show that Neural-Rendezvous provides a high probability exponential bound on the expected spacecraft delivery error, the proof of which leverages stochastic incremental stability analysis. In particular, it is used to construct a non-negative function with a supermartingale property, explicitly accounting for the ISO state uncertainty and the local nature of nonlinear state estimation guarantees. In numerical simulations, Neural-Rendezvous is demonstrated to satisfy the expected error bound for 100 ISO candidates. This performance is also empirically validated using our spacecraft simulator and in high-conflict and distributed unmanned aerial vehicle swarm reconfiguration with up to 20 unmanned aerial vehicles.

Shuttlecock Detection Using Residual Learning in U-Net Architecture

This paper introduces an enhanced approach for detecting shuttlecock. Detecting fast-moving objects, such as a shuttlecock, is crucial in various applications, including badminton video analysis and object tracking. Many deep-learning techniques have been proposed in literature to address this challenge. However, low image quality, motion blur, afterimage, and short-term occlusion can hinder accurate detection. To overcome these limitations, this research focuses on improving the existing method called TrackNetV2, which utilizes the U-Net architecture. The primary enhancement proposed in this study is incorporating residual learning within the U-Net architecture, emphasizing processing speed, prediction accuracy, and precision. Specifically, each U-Net layer is augmented with a residual layer, enhancing the network's overall performance. The results demonstrate that our proposed method outperforms the existing detection accuracy and reliability technique.

Top-down modulation of the retinal code via histaminergic neurons of the hypothalamus.

The mammalian retina is considered an autonomous circuit, yet work dating back to Ramon y Cajal indicates that it receives inputs from the brain. How such inputs affect retinal processing has remained unknown. We confirmed brain-to-retina projections of histaminergic neurons from the mouse hypothalamus. Histamine application ex vivo altered the activity of various retinal ganglion cells (RGCs), including direction-selective RGCs that gained responses to high motion velocities. These results were reproduced in vivo with optic tract recordings where histaminergic retinopetal axons were activated chemogenetically. Such changes could improve vision of fast-moving objects (e.g., while running), which fits with the known increased activity of histaminergic neurons during arousal. An antihistamine drug reduced optomotor responses to high-speed moving stimuli in freely moving mice. In humans, the same antihistamine nonuniformly modulated visual sensitivity across the visual field, indicating an evolutionary conserved function of the histaminergic system. Our findings expose a previously unappreciated role for brain-to-retina projections in modulating retinal function.

Enhanced 3D object detection for autonomous driving: A spatial-temporal alignment approach in Bird's Eye View scenarios

This paper presents a novel 3D object detection algorithm designed for Bird's Eye View (BEV) scenarios, which significantly improves detection capabilities by integrating spatial and temporal features. The core of our approach is the spatial-temporal alignment module that efficiently processes information across different time steps and spatial locations, enhancing the precision and robustness of object detection. We employ a temporal self-attention mechanism to capture the motion information of objects over time, allowing the model to correlate features across various time steps for identifying and tracking moving objects. Additionally, a spatial cross-attention mechanism is utilized to focus on spatial features within regions of interest, promoting interactions between features extracted from camera views and BEV queries. Our method also implements temporal feature integration and multi-scale feature fusion to enhance detection stability and accuracy for fast-moving objects and to capture multi-scale context information, respectively. The model employs an enriched feature set post alignment for 3D bounding box prediction, ascertaining the position, dimensions, and orientation of objects. We conducted experiments on two public datasets for autonomous driving nuScenes and Waymo Open Dataset, demonstrating that our method outperforms previous BEVFormer and other state-of-the-art methods in terms of detection accuracy and robustness. The paper concludes with potential future directions for optimizing the BEVFormer model's performance and exploring its application in broader scenarios and tasks.

Machine learning-based bee recognition and tracking for advancing insect behavior research

The study of insect behavior, particularly that of honey bees, has a broad scope and significance. Tracking bee flying patterns grants much helpful information about bee behavior. However, tracking a small yet fast-moving object, such as a bee, is difficult. Hence, we present artificial intelligence, machine-learning-based bee recognition, and tracking systems to assist the researcher in studying the bee’s behavior. To develop a machine learning system, a labeled database is required for model training. To address this, we implemented an automated system for analyzing and labeling bee videos. This labeled database served as the foundation for two distinct bee-tracking solutions. The first solution (planar bee tracking system) tracked individual bees in closed mazes using a neural network. The second solution (spatial bee tracking system) utilized a neural network and a tracking algorithm to identify and track flying bees in open environments. Both systems tackle the challenge of tracking small-bodied creatures with rapid and diverse movement patterns. Although we applied these systems to entomological cognition research in this paper, their relevance extends to general insect research and developing tracking solutions for small organisms with swift movements. We present the complete architecture and detailed methodologies to facilitate the utilization of these models in future research endeavors. Our approach is a simple and inexpensive method that contributes to the growing number of image-analysis tools used for tracking animal movement, with future potential applications under less sterile field conditions. The tools presented in this paper could assist the study of movement ecology, specifically in insects, by providing accurate movement specifications. Following the movement of pollinators or natural enemies, for example, greatly contributes to the study of pollination or biological control, respectively, in natural and agro-ecosystems.

Aperture photometry on asteroid trails. Detection of the fastest-rotating near-Earth object

Near-Earth objects (NEOs) on an impact course with Earth can move at high angular speeds. Understanding their properties, including their rotation state, is crucial for assessing impact risks and mitigation strategies. Traditional photometric methods face challenges in accurately collecting data on fast-moving NEOs. This study introduces an innovative approach to aperture photometry, tailored to analyzing trailed images of fast-moving NEOs. Our primary aim is to extract rotation state information for fast rotators. We applied our approach to the trailed images of three asteroids: 2023 CX1, 2024 BX1, and 2024 EF, which were either on a collision course or on a close fly-by with Earth, resulting in high angular velocities. By adjusting the aperture size, we controlled the effective instantaneous exposure time of the asteroid to increase the sampling rate of photometric variations. This enabled us to detect short rotation periods that would be challenging to derive with conventional methods. Our analysis shows that trailed photometry significantly reduces the overhead time associated with CCD readout, enhancing the sampling rate of the photometric variations. We demonstrate that this technique is particularly effective for fast-moving objects, providing reliable photometric data when the object is at its brightest and closest to Earth. For asteroid 2024 BX1, we detect a rotation period of $2.5888 0.0002$ seconds, the shortest ever recorded. We discuss under what circumstances it is most efficient to use trailed observations coupled with aperture photometry for studying the rotation characteristics of NEOs.

Prior-free 3D tracking of a fast-moving object at 6667 frames per second with single-pixel detectors.

Real-time tracking and 3D trajectory computation of fast-moving objects is a promising technology, especially in the field of autonomous driving. However, existing image-based tracking methods face significant challenges when it comes to real-time tracking, primarily due to the limitation of storage space and computational resources. Here, we propose a novel approach that enables real-time 3D tracking of a fast-moving object without any prior motion information and at a very low computational cost. To enable 3D coordinate synthesis with a space-efficient optical setup, geometric moment patterns are projected on two non-orthogonal planes with a spatial resolution of 125 μm. Our experiment demonstrates an impressive tracking speed of 6667 frames per second (FPS) with a 20 kHz digital micromirror device (DMD), which is more than 200 times faster than the widely adopted video-based tracking methods. To the best of our knowledge, this is the highest tracking speed record in the field of single-pixel 3D trajectory tracking. This method promotes the development of real-time tracking techniques with single-pixel imaging (SPI).

"I Go Outdoors for Activities Every Day": Go-Along With Seniors With Slow Walking Speeds to Explore Environmental Factors Influencing Mobility.

This study aims to: 1) Explore the mobility experiences of seniors with slow walking speeds (SSWS) in urban neighborhoods; and 2) Investigate their environmental barriers and supports. Go-along interviews were conducted with 36 SSWS residing in urban neighborhoods of Chongqing City, China. The mobility patterns and built environment factors influencing their mobility were revealed through cartographic analysis and thematic analysis. SSWS primarily focused their activities within a 400-meter radius of their homes. Built environment themes included topography, neighborhood services, sidewalks, seating, traffic safety, weather, greenery, and lighting. Significant mobility barriers included long stairs, steep slopes, fast-moving objects on sidewalks, road crossings, and fast traffic. Available handrails, nearby food-service places, ample seating, and greenery were identified as supportive factors for their mobility. This study stands out as the first to specifically examine the mobility of SSWS within the built environment. We suggest that SSWS should be taken into account when establishing a benchmark for general design frameworks. These improvements not only contribute to the mobility of slow walkers but also have positive impacts on the broader population.

A deep analysis for New Horizons’ KBO search images

Abstract Observation datasets acquired by the Hyper Suprime-Cam (HSC) on the Subaru Telescope for NASA’s New Horizons mission target search were analyzed through a method devised by JAXA. The method makes use of Field Programmable Gate arrays and was originally used to detect fast-moving objects such as space debris or near-Earth asteroids. Here we present an application of the method to detect slow-moving Kuiper Belt Objects (KBOs) in the New Horizons target search observations. A cadence that takes continuous images of one HSC field of view for half a night fits the method well. The observations for the New Horizons Kuiper Belt Extended Mission (NH/KEM) using HSC began in 2020 May, and are ongoing. Here we show our result of the analysis of the dataset acquired from 2020 May through 2021 June that have already passed the proprietary period and are open to the public. We detected 84 KBO candidates in the 2020 June and 2021 June datasets, when the observation field was close to opposition.

The oldest stars with low neutron-capture element abundances and origins in ancient dwarf galaxies

ABSTRACT We present a detailed chemical abundance and kinematic analysis of six extremely metal-poor (−4.2 ≤ [Fe/H] ≤−2.9) halo stars with very low neutron-capture abundances ([Sr/H] and [Ba/H]) based on high-resolution Magellan/MIKE spectra. Three of our stars have [Sr/Ba] and [Sr/H] ratios that resemble those of metal-poor stars in ultra-faint dwarf galaxies (UFDs). Since early UFDs may be the building blocks of the Milky Way, extremely metal-poor halo stars with low, UFD-like Sr and Ba abundances may thus be ancient stars from the earliest small galactic systems that were accreted by the proto-Milky Way. We label these objects as Small Accreted Stellar System (SASS) stars, and we find an additional 61 similar ones in the literature. A kinematic analysis of our sample and literature stars reveals them to be fast-moving halo objects, all with retrograde motion, indicating an accretion origin. Because SASS stars are much brighter than typical UFD stars, identifying them offers promising ways towards detailed studies of early star formation environments. From the chemical abundances of SASS stars, it appears that the earliest accreted systems were likely enriched by a few supernovae whose light element yields varied from system to system. Neutron-capture elements were sparsely produced and/or diluted, with r-process nucleosynthesis playing a role. These insights offer a glimpse into the early formation of the Galaxy. Using neutron-capture elements as a distinguishing criterion for early formation, we have access to a unique metal-poor population that consists of the oldest stars in the universe.

Dual-domain mean-reverting diffusion model-enhanced temporal compressive coherent diffraction imaging

Temporal compressive coherent diffraction imaging is a lensless imaging technique with the capability to capture fast-moving small objects. However, the accuracy of imaging reconstruction is often hindered by the loss of frequency domain information, a critical factor limiting the quality of the reconstructed images. To improve the quality of these reconstructed images, a method dual-domain mean-reverting diffusion model-enhanced temporal compressive coherent diffraction imaging (DMDTC) has been introduced. DMDTC leverages the mean-reverting diffusion model to acquire prior information in both frequency and spatial domain through sample learning. The frequency domain mean-reverting diffusion model is employed to recover missing information, while hybrid input-output algorithm is carried out to reconstruct the spatial domain image. The spatial domain mean-reverting diffusion model is utilized for denoising and image restoration. DMDTC has demonstrated a significant enhancement in the quality of the reconstructed images. The results indicate that the structural similarity and peak signal-to-noise ratio of images reconstructed by DMDTC surpass those obtained through conventional methods. DMDTC enables high temporal frame rates and high spatial resolution in coherent diffraction imaging.

Recognizing Ultra-High-Speed Moving Objects with Bio-Inspired Spike Camera

Bio-inspired spike camera mimics the sampling principle of primate fovea. It presents high temporal resolution and dynamic range, showing great promise in fast-moving object recognition. However, the physical limit of CMOS technology in spike cameras still hinders their capability of recognizing ultra-high-speed moving objects, e.g., extremely fast motions cause blur during the imaging process of spike cameras. This paper presents the first theoretical analysis for the causes of spiking motion blur and proposes a robust representation that addresses this issue through temporal-spatial context learning. The proposed method leverages multi-span feature aggregation to capture temporal cues and employs residual deformable convolution to model spatial correlation among neighbouring pixels. Additionally, this paper contributes an original real-captured spiking recognition dataset consisting of 12,000 ultra-high-speed (equivalent speed > 500 km/h) moving objects. Experimental results show that the proposed method achieves 73.2% accuracy in recognizing 10 classes of ultra-high-speed moving objects, outperforming all existing spike-based recognition methods. Resources will be available at https://github.com/Evin-X/UHSR.

Error Analysis for Rotating-drift-scan Charge-coupled Device Observation of Near-Earth Asteroids

The apparent velocities of near-Earth asteroids (NEAs) are usually high when they pass by Earth. Observing these fast-moving objects with long exposure times would cause their images to streak and significantly decrease the precision of astronomical measurements. The rotating-drift-scan (RDS) charge-coupled device technique is a promising approach to observe fast-moving NEAs during their close approaches to Earth. By rotating the camera of a telescope, an NEA can be observed in the time delay integration mode. This allows the asteroid to be imaged as a point source, even with a long exposure time. Here, we thoroughly present the RDS follow-up observation and orbit determination of a newly discovered NEA 2023 BJ7. This technique makes an impactful contribution to improving the NEA's orbit accuracy by extending the observation arc. A detailed statistical analysis of the astrometric error was conducted, revealing that RDS observations can achieve a competitive accuracy with an rms error of 0.″24 in right ascension and 0.″32 in declination. The instability of the telescope is thought to be the main reason affecting the internal precision. Furthermore, the RDS technique excels at observing fast-moving NEAs, as well as newly discovered NEAs without accurate ephemerides. For NEAs with rates of motion exceeding 10 deg day−1, the rms of RDS observation residuals is 0.″35 in the along-track direction and 0.″23 in the cross-track. With this technique, a network of small-aperture telescopes would substantially benefit our global NEAs monitoring system to ensure Earth’s safety from any asteroid impacts.

Near-Earth Object Observations using Synthetic Tracking

Synthetic tracking (ST) has emerged as a potent technique for observing fast-moving near-Earth objects (NEOs), offering enhanced detection sensitivity and astrometric accuracy by avoiding trailing loss. This approach also empowers small telescopes to use prolonged integration times to achieve high sensitivity for NEO surveys and follow-up observations. In this study, we present the outcomes of ST observations conducted with Pomona College’s 1 m telescope at the Table Mountain Facility and JPL’s robotic telescopes at the Sierra Remote Observatory. The results showcase astrometric accuracy statistics comparable to stellar astrometry, irrespective of an object’s rate of motion, and the capability to detect faint asteroids beyond 20.5th magnitude using 11 inch telescopes. Furthermore, we detail the technical aspects of data processing, including the correction of differential chromatic refraction in the atmosphere and accurate timing for image stacking, which contribute to achieving precise astrometry. We also provide compelling examples that showcase the robustness of ST even when asteroids closely approach stars or bright satellites cause disturbances. Moreover, we illustrate the proficiency of ST in recovering NEO candidates with highly uncertain ephemerides. As a glimpse of the potential of NEO surveys utilizing small robotic telescopes with ST, we present significant statistics from our NEO survey conducted for testing purposes. These findings underscore the promise and effectiveness of ST as a powerful tool for observing fast-moving NEOs, offering valuable insights into their trajectories and characteristics. Overall, the adoption of ST stands to revolutionize fast-moving NEO observations for planetary defense and studying these celestial bodies.

A Bidirectional Scoring Strategy-Based Transformation Matrix Estimation of Dynamic Factors in Environmental Sensing

Simultaneous localization and mapping (SLAM) is the technological basis of environmental sensing, and it has been widely applied to autonomous navigation. In combination with deep learning methods, dynamic SLAM algorithms have emerged to provide a certain stability and accuracy in dynamic scenes. However, the robustness and accuracy of existing dynamic SLAM algorithms are relatively low in dynamic scenes, and their performance is affected by potential dynamic objects and fast-moving dynamic objects. To solve the positioning interference caused by these dynamic objects, this study proposes a geometric constraint algorithm that utilizes a bidirectional scoring strategy for the estimation of a transformation matrix. First, a geometric constraint function is defined according to the Euclidean distance between corresponding feature points and the average distance of the corresponding edges. This function serves as the basis for determining abnormal scores for feature points. By utilizing these abnormal score values, the system can identify and eliminate highly dynamic feature points. Then, a transformation matrix estimation based on the filtered feature points is adopted to remove more outliers, and a function for evaluating the similarity of key points in two images is optimized during this process. Experiments were performed based on the TUM dynamic target dataset and Bonn RGB-D dynamic dataset, and the results showed that the added dynamic detection method effectively improved the performance compared to state of the art in highly dynamic scenarios.

Real-Time Moving Object Tracking on Smartphone Using Cradle Head Servo Motor.

The increasing demand for artificially intelligent smartphone cradles has prompted the need for real-time moving object detection. Real-time moving object tracking requires the development of algorithms for instant tracking analysis without delays. In particular, developing a system for smartphones should consider different operating systems and software development environments. Issues in current real-time moving object tracking systems arise when small and large objects coexist, causing the algorithm to prioritize larger objects or struggle with consistent tracking across varying scales. Fast object motion further complicates accurate tracking and leads to potential errors and misidentification. To address these issues, we propose a deep learning-based real-time moving object tracking system which provides an accuracy priority mode and a speed priority mode. The accuracy priority mode achieves a balance between the high accuracy and speed required in the smartphone environment. The speed priority mode optimizes the speed of inference to track fast-moving objects. The accuracy priority mode incorporates CSPNet with ResNet to maintain high accuracy, whereas the speed priority mode simplifies the complexity of the convolutional layer while maintaining accuracy. In our experiments, we evaluated both modes in terms of accuracy and speed.

A Semantic Tree-Based Fast-Moving Object Trajectory Tracking Algorithm for Table Tennis

Table tennis is a popular sport around the world. A key technology in table tennis education and analysis system is reconstructing the trajectory of the fast-moving ball from videos. Typically the table tennis ball is too small and barely visible in the video, making it difficult to be recognized directly by detection models like YOLO. However, table tennis balls usually has obvious motion features, which are usually not found in similar false targets. It inspired the authors to first find all candidate targets and then use the motion features of table tennis ball to select them out. In this article, the authors propose a tree-based algorithm named T-FORT to track the ball and reconstruct its trajectory. Specifically, they consider all the possible objects in a tree-framework, and identify the real target by integrating visual features and moving patterns. The authors conduct a set of experiments on three datasets to evaluate the effectiveness and performance of the proposed algorithm. The experimental results show that the proposed method is more precise than existing algorithms, and is robust in various scenarios.

Fuzzy-based video compression using bilinear fuzzy relation equations

We present a novel color video compression method using the greatest solution of a system of bilinear fuzzy relation equations to assess the similarity between frames. The frames in each band are treated separately and each frame is classified as an Intra frame or a Predictive frame. A frame is labelled as Predictive frame, and compressed more than an Intra-frame, if the similarity value with the previous Intra frame is higher than a selected threshold; A pre-processing activity is performed to select the optimal threshold value of the similarity between frames. The proposed method allows to supply a high quality of the reconstructed frames and has the advantage of not requiring high CPU time and memory storage for its execution; it was tested on color videos of the Fast-Moving Objects dataset; the results show that it produces better performances than the Lukasiewicz similarity-based video compression method and comparable with those achieved by MPEG-4 and the deep learning video compression method DVC_pro. The results show that the quality of the reconstructed frames obtained with BFRE is comparable with that of DVC Pro, but has a lower computational complexity, providing better performances in terms of video encoding speed.

A Biomimetic Visual Detection Model: Event-Driven LGMDs Implemented With Fractional Spiking Neuron Circuits.

Lobula giant motion detectors (LGMDs) in locusts effectively predict collisions and trigger avoidance, with potential applications in autonomous driving and UAVs. Research on LGMD characteristics splits into two views: one focusing on the presynaptic visual pathway, the other on the postsynaptic LGMD neurons. Both perspectives have support, leading to two computational models, but they lack a biophysical description of the individual LGMD neuron behavior. This paper aims to mimic and explain LGMD behavior based on fractional spiking neurons (FSNs) and construct a biomimetic visual model for the LGMD compatible with these characteristics. We implement the visual model using an event camera to simulate photoreceptors and follow the ON/OFF visual pathway, incorporating lateral inhibition to mimic the LGMD system from the bottom up. Second, most computational models of motion perception use only the dendrites within the LGMD neurons as the ideal pathway for linear summation, ignoring dendritic effects inducing neuronal properties. Thus, we introduce FSN circuits by altering dendritic morphological parameters to simulate multi-scale spike frequency adaptation (SFA) observed in LGMDs. Additionally, we add one more circuit of dendritic trees into the FSNs to be compatible with the postsynaptic feed-forward inhibition (FFI) in LGMD neurons, providing a novel explanatory and predictive model. We test that the event-driven biomimetic visual model can achieve collision detection and looming selection in different complex scenes, especially fast-moving objects.

The Orthogonal Time Frequency Space (OTFS) Technique In 6G Wireless Communications

The orthogonal time-frequency space (OTFS) technique is a potential waveform modulation method that modulates data in the delaydoppler (DD) domain. OTFS differs from traditional multiplexing techniques by utilizing two-dimensional modulation to switch between the time-frequency (TF) domain and the delay-Doppler domain. This allows for handling Doppler shifts caused by fast-moving objects, a capability lacking in traditional modulation techniques like orthogonal frequency division multiplexing (OFDM). The primary goal of this paper is to offer an overview and short survey of this new topic, highlighting its system model. We also examine key aspects of OTFS modulation such as data detection methods, channel estimation, MIMO, and multiuser systems.