Search

Vision-based motion estimation methods show promise in accurately and unobtrusively estimating human body motion for healthcare purposes. However, these methods are not specifically designed for healthcare purposes and face challenges in real-world applications. Human pose estimation methods often lack the accuracy needed for detecting fine-grained, subtle body movements, while optical flow-based methods struggle with poor lighting conditions and unseen real-world data. These issues result in human body motion estimation errors, particularly during critical medical situations where the body is motionless, such as during unconsciousness. To address these challenges and improve the accuracy of human body motion estimation for healthcare purposes, we propose the OPPH operator designed to enhance current vision-based motion estimation methods. This operator, which considers human body movement and noise properties, functions as a multi-stage filter. Results tested on two real-world and one synthetic human motion dataset demonstrate that the operator effectively removes real-world noise, significantly enhances the detection of motionless states, maintains the accuracy of estimating active body movements, and maintains long-term body movement trends. This method could be beneficial for analyzing both critical medical events and chronic medical conditions.

Aging and chronic conditions affect older adults' daily lives, making early detection of developing health issues crucial. Weakness, common in many conditions, alters physical movements and daily activities subtly. However, detecting such changes can be challenging due to their subtle and gradual nature. To address this, we employ a non-intrusive camera sensor to monitor individuals' daily sitting and relaxing activities for signs of weakness. We simulate weakness in healthy subjects by having them perform physical exercise and observing the behavioral changes in their daily activities before and after workouts. The proposed system captures fine-grained features related to body motion, inactivity, and environmental context in real-time while prioritizing privacy. A Bayesian Network is used to model the relationships between features, activities, and health conditions. We aim to identify specific features and activities that indicate such changes and determine the most suitable time scale for observing the change. Results show 0.97 accuracy in distinguishing simulated weakness at the daily level. Fine-grained behavioral features, including non-dominant upper body motion speed and scale, and inactivity distribution, along with a 300-second window, are found most effective. However, individual-specific models are recommended as no universal set of optimal features and activities was identified across all participants.

Suzaku X-ray observations of the Type Ia supernova remnant (SNR) 3C 397 discovered exceptionally high mass ratios of Mn/Fe, Ni/Fe, and Cr/Fe, consistent with a near $M_{\rm Ch}$ progenitor white dwarf (WD). The Suzaku observations have established 3C 397 as our best candidate for a near-$M_{\rm Ch}$ SNR Ia, and opened the way to address additional outstanding questions about the origin and explosion mechanism of these transients. In particular, subsequent XMM-Newton observations revealed an unusually clumpy distribution of iron group elemental (IGE) abundances within the ejecta of 3C 397. In this paper, we undertake a suite of two dimensional hydrodynamical models, varying both the explosion mechanism -- either deflagration-to-detonation (DDT), or pure deflagration -- WD progenitors, and WD progenitor metallicity, and analyze their detailed nucleosynthetic abundances and associated clumping. We find that pure deflagrations naturally give rise to clumpy distributions of neutronized species concentrated towards the outer limb of the remnant, and confirm DDTs have smoothly structured ejecta with a central concentration of neutronization. Our findings indicate that 3C 397 was most likely a pure deflagration of a high central density WD. We discuss a range of implications of these findings for the broader SN Ia progenitor problem., Comment: 13 pages, 6 figures, Stable mean nucleosynthetic yields' datasets are available at https://doi.org/10.5281/zenodo.10927265 . Comments are welcome

Three-dimensional data registration is an established yet challenging problem that is key in many different applications, such as mapping the environment for autonomous vehicles, and modeling objects and people for avatar creation, among many others. Registration refers to the process of mapping multiple data into the same coordinate system by means of matching correspondences and transformation estimation. Novel proposals exploit the benefits of deep learning architectures for this purpose, as they learn the best features for the data, providing better matches and hence results. However, the state of the art is usually focused on cases of relatively small transformations, although in certain applications and in a real and practical environment, large transformations are very common. In this paper, we present ReLaTo (Registration for Large Transformations), an architecture that faces the cases where large transformations happen while maintaining good performance for local transformations. This proposal uses a novel Softmax pooling layer to find correspondences in a bilateral consensus manner between two point sets, sampling the most confident matches. These matches are used to estimate a coarse and global registration using weighted Singular Value Decomposition (SVD). A target-guided denoising step is then applied to both the obtained matches and latent features, estimating the final fine registration considering the local geometry. All these steps are carried out following an end-to-end approach, which has been shown to improve 10 state-of-the-art registration methods in two datasets commonly used for this task (ModelNet40 and KITTI), especially in the case of large transformations.

Despite recent advances in video action recognition achieving strong performance on existing benchmarks, these models often lack robustness when faced with natural distribution shifts between training and test data. We propose two novel evaluation methods to assess model resilience to such distribution disparity. One method uses two different datasets collected from different sources and uses one for training and validation, and the other for testing. More precisely, we created dataset splits of HMDB-51 or UCF-101 for training, and Kinetics-400 for testing, using the subset of the classes that are overlapping in both train and test datasets. The other proposed method extracts the feature mean of each class from the target evaluation dataset's training data (i.e. class prototype) and estimates test video prediction as a cosine similarity score between each sample to the class prototypes of each target class. This procedure does not alter model weights using the target dataset and it does not require aligning overlapping classes of two different datasets, thus is a very efficient method to test the model robustness to distribution shifts without prior knowledge of the target distribution. We address the robustness problem by adversarial augmentation training - generating augmented views of videos that are "hard" for the classification model by applying gradient ascent on the augmentation parameters - as well as "curriculum" scheduling the strength of the video augmentations. We experimentally demonstrate the superior performance of the proposed adversarial augmentation approach over baselines across three state-of-the-art action recognition models - TSM, Video Swin Transformer, and Uniformer. The presented work provides critical insight into model robustness to distribution shifts and presents effective techniques to enhance video action recognition performance in a real-world deployment.

Deformable linear object (DLO) manipulation is needed in many fields. Previous research on deformable linear object (DLO) manipulation has primarily involved parallel jaw gripper manipulation with fixed grasping positions. However, the potential for dexterous manipulation of DLOs using an anthropomorphic hand is under-explored. We present DexDLO, a model-free framework that learns dexterous dynamic manipulation policies for deformable linear objects with a fixed-base dexterous hand in an end-to-end way. By abstracting several common DLO manipulation tasks into goal-conditioned tasks, our DexDLO can perform these tasks, such as DLO grabbing, DLO pulling, DLO end-tip position controlling, etc. Using the Mujoco physics simulator, we demonstrate that our framework can efficiently and effectively learn five different DLO manipulation tasks with the same framework parameters. We further provide a thorough analysis of learned policies, reward functions, and reduced observations for a comprehensive understanding of the framework., Comment: 7 pages, 8 figures, submitted to ICRA 2024

Type Ia supernovae (SNe Ia) are standardizable cosmological candles which led to the discovery of the accelerating universe. However, the physics of how white dwarfs (WDs) explode and lead to SNe Ia is still poorly understood. The initiation of the detonation front which rapidly disrupts the WD is a crucial element of the puzzle, and global 3D simulations of SNe Ia cannot resolve the requisite length scales to capture detonation initiation. In this work, we elucidate a theoretical criterion for detonation initiation in the distributed burning regime. We test this criterion against local 3D driven turbulent hydrodynamical simulations within electron-degenerate WD matter consisting initially of pure helium. We demonstrate a novel pathway for detonation, in which strong turbulent dissipation rapidly heats the helium, and forms carbon nuclei sufficient to lead to a detonation through accelerated burning via $\alpha$ captures. Simulations of strongly-driven turbulent conditions lead to detonations at a mean density of $10^6$ g cm$^{-3}$ and mean temperature of $1.4 - 1.8 \times 10^9$ K, but fail to detonate at a lower density of $10^5$ g cm$^{-3}$, in excellent agreement with theoretical predictions., Comment: 10 pages, 3 figures. Submitted to ApJ. Comments welcome

A new application for real-time monitoring of the lack of movement in older adults' own homes is proposed, aiming to support people's lives and independence in their later years. A lightweight camera monitoring system, based on an RGB-D camera and a compact computer processor, was developed and piloted in community homes to observe the daily behavior of older adults. Instances of body inactivity were detected in everyday scenarios anonymously and unobtrusively. These events can be explained at a higher level, such as a loss of consciousness or physiological deterioration. The accuracy of the inactivity monitoring system is assessed, and statistics of inactivity events related to the daily behavior of older adults are provided. The results demonstrate that our method achieves high accuracy in inactivity detection across various environments and camera views. It outperforms existing state-of-the-art vision-based models in challenging conditions like dim room lighting and TV flickering. However, the proposed method does require some ambient light to function effectively., Comment: 10 pages, 11 figures, paper

Recovering an outdoor environment's surface mesh is vital for an agricultural robot during task planning and remote visualization. Our proposed solution is based on a newly-designed panoramic stereo camera along with a hybrid novel software framework that consists of three fusion modules. The panoramic stereo camera with a pentagon shape consists of 5 stereo vision camera pairs to stream synchronized panoramic stereo images for the following three fusion modules. In the disparity fusion module, rectified stereo images produce the initial disparity maps using multiple stereo vision algorithms. Then, these initial disparity maps, along with the intensity images, are input into a disparity fusion network to produce refined disparity maps. Next, the refined disparity maps are converted into full-view point clouds or single-view point clouds for the pose fusion module. The pose fusion module adopts a two-stage global-coarse-to-local-fine strategy. In the first stage, each pair of full-view point clouds is registered by a global point cloud matching algorithm to estimate the transformation for a global pose graph's edge, which effectively implements loop closure. In the second stage, a local point cloud matching algorithm is used to match single-view point clouds in different nodes. Next, we locally refine the poses of all corresponding edges in the global pose graph using three proposed rules, thus constructing a refined pose graph. The refined pose graph is optimized to produce a global pose trajectory for volumetric fusion. In the volumetric fusion module, the global poses of all the nodes are used to integrate the single-view point clouds into the volume to produce the mesh of the whole garden. The proposed framework and its three fusion modules are tested on a real outdoor garden dataset to show the superiority of the performance., Comment: 32 pages

To enable a mobile manipulator to perform human tasks from a single teaching demonstration is vital to flexible manufacturing. We call our proposed method MMPA (Mobile Manipulator Process Automation with One-shot Teaching). Currently, there is no effective and robust MMPA framework which is not influenced by harsh industrial environments and the mobile base's parking precision. The proposed MMPA framework consists of two stages: collecting data (mobile base's location, environment information, end-effector's path) in the teaching stage for robot learning; letting the end-effector repeat the nearly same path as the reference path in the world frame to reproduce the work in the automation stage. More specifically, in the automation stage, the robot navigates to the specified location without the need of a precise parking. Then, based on colored point cloud registration, the proposed IPE (Iterative Pose Estimation by Eye & Hand) algorithm could estimate the accurate 6D relative parking pose of the robot arm base without the need of any marker. Finally, the robot could learn the error compensation from the parking pose's bias to modify the end-effector's path to make it repeat a nearly same path in the world coordinate system as recorded in the teaching stage. Hundreds of trials have been conducted with a real mobile manipulator to show the superior robustness of the system and the accuracy of the process automation regardless of the harsh industrial conditions and parking precision. For the released code, please contact marketing@amigaga.com, Comment: 25 pages

The radiosotope $^{44}$Ti is produced through $\alpha$-rich freezeout and explosive helium burning in type Ia supernovae (SNe Ia). In this paper, we discuss how the detection of $^{44}$Ti, either through late-time light curves of SNe Ia, or directly via gamma rays, can uniquely constrain the origin of SNe Ia. In particular, building upon recent advances in the hydrodynamical simulation of helium-ignited double white dwarf binaries, we demonstrate that the detection of $^{44}$Ti in a nearby SN Ia or in a young galactic supernova remnant (SNR) can discriminate between the double-detonation and double-degenerate channels of sub-Chandrasekhar (sub-$M_{\rm Ch}$) and near-Chandrasekhar (near-$M_{\rm Ch}$) SNe Ia. In addition, we predict that the late-time light curves of calcium-rich transients are entirely dominated by $^{44}$Ti.

Type Ia supernovae (SNe Ia) play a crucial role as standardizable candles in measurements of the Hubble constant and dark energy. Increasing evidence points towards multiple possible explosion channels as the origin of normal SNe Ia, with possible systematic effects on the determination of cosmological parameters. We present, for the first time, a comprehensive comparison of publicly-available SN Ia model nucleosynthetic data with observations of late-time light curve observations of SN Ia events. These models span a wide range of white dwarf (WD) progenitor masses, metallicities, explosion channels, and numerical methodologies. We focus on the influence of $^{57}$Ni and its isobaric decay product $^{57}$Co in powering the late-time ($t > 1000$ d) light curves of SNe Ia. $^{57}$Ni and $^{57}$Co are neutron-rich relative to the more abundant radioisotope $^{56}$Ni, and are consequently a sensitive probe of neutronization at the higher densities of near-Chandrashekhar (near-$M_{\rm Ch}$) progenitor WDs. We demonstrate that observations of one SN Ia event, SN 2015F is only consistent with a sub-$M_{\rm Ch}$ WD progenitor. Observations of four other events (SN 2011fe, SN 2012cg, SN 2014J, SN2013aa) are consistent with both near-$M_{\rm Ch}$ and sub-$M_{\rm Ch}$ progenitors. Continued observations of late-time light curves of nearby SNe Ia will provide crucial information on the nature of the SN Ia progenitors., Comment: 13 pages, 5 figures, 3 tables. Accepted for publication in MNRAS. Comments welcome!

The origins of type Ia supernovae (SNe Ia) are still debated. Some of the leading scenarios involve a double detonation in double white dwarf (WD) systems. In these scenarios, helium shell detonation occurs on top of a carbon-oxygen (CO) WD, which then drives the detonation of the CO-core, producing a SN Ia. Extensive studies have been done on the possibility of a double helium detonation, following a dynamical helium mass-transfer phase onto a CO-WD. However, 3D self-consistent modeling of the double-WD system, the mass transfer, and the helium shell detonation have been little studied. Here we use 3D hydrodynamical simulations to explore this case in which a helium detonation occurs near the point of Roche lobe overflow of the donor WD and may lead to an SN Ia through the dynamically driven double-degenerate double-detonation (D6) mechanism. We find that the helium layer of the accreting primary WD does undergo a detonation, while the underlying carbon-oxygen core does not, leading to an extremely rapid and faint nova-like transient instead of a luminous SN Ia event. This failed core detonation suggests that D6 SNe Ia may be restricted to the most massive carbon-oxygen primary WDs. We highlight the nucleosynthesis of the long-lived radioisotope $^{44}$Ti during explosive helium burning, which may serve as a hallmark both of successful as well as failed D6 events which subsequently detonate as classical double-degenerate mergers., Comment: 11 pages, 4 figures, 1 table. Published in the Astrophysical Journal Letters

Recent observational evidence has demonstrated that white dwarf (WD) mergers are a highly efficient mechanism for mass accretion onto WDs in the galaxy. In this paper, we show that WD mergers naturally produce highly-magnetized, uniformly-rotating WDs, including a substantial population within a narrow mass range close to the Chandrasekhar mass ($M_{\rm Ch}$). These near-$M_{\rm Ch}$ WD mergers subsequently undergo rapid spin up and compression on a $\sim 10^2$ yr timescale, either leading to central ignition and a normal SN Ia via the DDT mechanism, or alternatively to a failed detonation and SN Iax through pure deflagration. The resulting SNe Ia and SNe Iax will have spectra, light curves, polarimetry, and nucleosynthetic yields similar to those predicted to arise through the canonical near-$M_{\rm Ch}$ single degenerate (SD) channel, but with a $t^{-1}$ delay time distribution characteristic of the double-degenerate (DD) channel. Furthermore, in contrast to the SD channel, WD merger near-$M_{\rm Ch}$ SNe Ia and SNe Iax will not produce observable companion signatures. We discuss a range of implications of these findings, from SNe Ia explosion mechanisms, to galactic nucleosynthesis of iron peak elements including manganese., Comment: 20 pages, 4 figures, nucleosynthetic yields included as ancillary data files. Accepted to The Astrophysical Journal. Comments welcome

The lip is a dominant dynamic facial unit when a person is speaking. Detecting lip events is beneficial to speech analysis and support for the hearing impaired. This paper proposes a 3D lip event detection pipeline that automatically determines the lip events from a 3D speaking lip sequence. We define a motion divergence measure using 3D lip landmarks to quantify the interframe dynamics of a 3D speaking lip. Then, we cast the interframe motion detection in a multi-temporal-resolution framework that allows the detection to be applicable to different speaking speeds. The experiments on the S3DFM Dataset investigate the overall 3D lip dynamics based on the proposed motion divergence. The proposed 3D pipeline is able to detect opening and closing lip events across 100 sequences, achieving a state-of-the-art performance.

Learning object-centric representations of multi-object scenes is a promising approach towards machine intelligence, facilitating high-level reasoning and control from visual sensory data. However, current approaches for unsupervised object-centric scene representation are incapable of aggregating information from multiple observations of a scene. As a result, these "single-view" methods form their representations of a 3D scene based only on a single 2D observation (view). Naturally, this leads to several inaccuracies, with these methods falling victim to single-view spatial ambiguities. To address this, we propose The Multi-View and Multi-Object Network (MulMON) -- a method for learning accurate, object-centric representations of multi-object scenes by leveraging multiple views. In order to sidestep the main technical difficulty of the multi-object-multi-view scenario -- maintaining object correspondences across views -- MulMON iteratively updates the latent object representations for a scene over multiple views. To ensure that these iterative updates do indeed aggregate spatial information to form a complete 3D scene understanding, MulMON is asked to predict the appearance of the scene from novel viewpoints during training. Through experiments, we show that MulMON better-resolves spatial ambiguities than single-view methods -- learning more accurate and disentangled object representations -- and also achieves new functionality in predicting object segmentations for novel viewpoints., Comment: Accepted at NeurIPS 2020 (Spotlight)

Learning object-centric scene representations is essential for attaining structural understanding and abstraction of complex scenes. Yet, as current approaches for unsupervised object-centric representation learning are built upon either a stationary observer assumption or a static scene assumption, they often: i) suffer single-view spatial ambiguities, or ii) infer incorrectly or inaccurately object representations from dynamic scenes. To address this, we propose Dynamics-aware Multi-Object Network (DyMON), a method that broadens the scope of multi-view object-centric representation learning to dynamic scenes. We train DyMON on multi-view-dynamic-scene data and show that DyMON learns -- without supervision -- to factorize the entangled effects of observer motions and scene object dynamics from a sequence of observations, and constructs scene object spatial representations suitable for rendering at arbitrary times (querying across time) and from arbitrary viewpoints (querying across space). We also show that the factorized scene representations (w.r.t. objects) support querying about a single object by space and time independently., Comment: Accepted at NeurIPS 2021

We present an innovative two-headed attention layer that combines geometric and latent features to segment a 3D scene into semantically meaningful subsets. Each head combines local and global information, using either the geometric or latent features, of a neighborhood of points and uses this information to learn better local relationships. This Geometric-Latent attention layer (Ge-Latto) is combined with a sub-sampling strategy to capture global features. Our method is invariant to permutation thanks to the use of shared-MLP layers, and it can also be used with point clouds with varying densities because the local attention layer does not depend on the neighbor order. Our proposal is simple yet robust, which allows it to achieve competitive results in the ShapeNetPart and ModelNet40 datasets, and the state-of-the-art when segmenting the complex dataset S3DIS, with 69.2% IoU on Area 5, and 89.7% overall accuracy using K-fold cross-validation on the 6 areas., Comment: Accepted in BMVC 2021

The supernova remnant (SNR) 3C 397 is thought to originate from a Type Ia supernova (SN Ia) explosion of a near-Chandrasekhar-mass ($M_{\rm Ch}$) progenitor, based on the enhanced abundances of Mn and Ni revealed by previous X-ray study with Suzaku. Here we report follow-up XMM-Newton observations of this SNR, conducted with the aim of investigating the detailed spatial distribution of the Fe-peak elements. We have discovered an ejecta clump with extremely high abundances of Ti and Cr, in addition to Mn, Fe, and Ni, in the southern part of the SNR. The Fe mass of this ejecta clump is estimated to be $\sim$ 0.06 $M_{\odot}$, under the assumption of a typical Fe yield for SNe Ia (i.e., $\sim$ 0.8 $M_{\odot}$). The observed mass ratios among the Fe-peak elements and Ti require substantial neutronization that is achieved only in the innermost regions of a near-$M_{\rm Ch}$ SN Ia with a central density of $\rho_c \sim 5 \times 10^9$ g cm$^{-3}$, significantly higher than typically assumed for standard near-$M_{\rm Ch}$ SNe Ia ($\rho_c \sim 2 \times 10^9$ g cm$^{-3}$). The overproduction of the neutron-rich isotopes (e.g., $^{50}$Ti and $^{54}$Cr) is significant in such high-$\rho_c$ SNe Ia, with respect to the solar composition. Therefore, if 3C 397 is a typical high-$\rho_c$ near-$M_{\rm Ch}$ SN Ia remnant, the solar abundances of these isotopes could be reproduced by the mixture of the high- and low-$\rho_c$ near-$M_{\rm Ch}$ and sub-$M_{\rm Ch}$ Type Ia events, with $\lesssim$ 20 % being high-$\rho_c$ near-$M_{\rm Ch}$., Comment: Accepted by ApJL; 9 pages with 4 figures and 1 table