The Gaia M dwarf gap, also known as the Jao Gap, is a novel feature discovered in the Gaia Data Release 2
This content will become publicly available on August 1, 2024
Open Anomalous Trajectory Recognition via Probabilistic Metric Learning
Typically, trajectories considered anomalous are the ones deviating from usual (e.g., traffic-dictated) driving patterns. However, this closed-set context fails to recognize the unknown anomalous trajectories, resulting in an insufficient self-motivated learning paradigm. In this study, we investigate the novel Anomalous Trajectory Recognition problem in an Open-world scenario (ATRO) and introduce a novel probabilistic Metric learning model, namely ATROM, to address it. Specifically, ATROM can detect the presence of unknown anomalous behavior in addition to identifying known behavior. It has a Mutual Interaction Distillation that uses contrastive metric learning to explore the interactive semantics regarding the diverse behavioral intents and a Probabilistic Trajectory Embedding that forces the trajectories with distinct behaviors to follow different Gaussian priors. More importantly, ATROM offers a probabilistic metric rule to discriminate between known and unknown behavioral patterns by taking advantage of the approximation of multiple priors. Experimental results on two large-scale trajectory datasets demonstrate the superiority of ATROM in addressing both known and unknown anomalous patterns.
more » « less- Award ID(s):
- 2030249
- NSF-PAR ID:
- 10488101
- Publisher / Repository:
- International Joint Conferences on Artificial Intelligence Organization
- Date Published:
- Journal Name:
- IJCAI
- ISSN:
- 1045-0823
- Page Range / eLocation ID:
- 2095 to 2103
- Format(s):
- Medium: X
- Location:
- Macau, SAR China
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract G versus BP-RP color–magnitude diagram. This gap represents a 17% decrease in stellar density in a thin magnitude band around the convective transition mass (∼0.35M ⊙) on the main sequence. Previous work has demonstrated a paucity of Hα emission coincident with theG magnitude of the Jao Gap in the solar neighborhood. The exact mechanism that results in this paucity is as of yet unknown; however, the authors of the originating paper suggest that it may be the result of complex variations to a star’s magnetic topology driven by the Jao Gap’s characteristic formation and breakdown of stars’ radiative transition zones. We present a follow-up investigating another widely used magnetic activity metric, Calciumii H and K emission. Caii H and K activity appears to share a similar anomalous behavior as Hα does near the Jao Gap magnitude. We observe an increase in star-to-star variation of magnetic activity near the Jao Gap. We present a toy model of a star’s magnetic field evolution, which demonstrates that this increase may be due to stochastic disruptions to the magnetic field originating from the periodic-mixing events characteristic of the convective kissing instabilities that drive the formation of the Jao Gap. -
In this paper, we develop a novel and safe control design approach that takes demonstrations provided by a human teacher to enable a robot to accomplish complex manipulation scenarios in dynamic environments. First, an overall task is divided into multiple simpler subtasks that are more appropriate for learning and control objectives. Then, by collecting human demonstrations, the subtasks that require robot movement are modeled by probabilistic movement primitives (ProMPs). We also study two strategies for modifying the ProMPs to avoid collisions with environmental obstacles. Finally, we introduce a rule-base control technique by utilizing a finite-state machine along with a unique means of control design for ProMPs. For the ProMP controller, we propose control barrier and Lyapunov functions to guide the system along a trajectory within the distribution defined by a ProMP while guaranteeing that the system state never leaves more than a desired distance from the distribution mean. This allows for better performance on nonlinear systems and offers solid stability and known bounds on the system state. A series of simulations and experimental studies demonstrate the efficacy of our approach and show that it can run in real time. Note to Practitioners —This paper is motivated by the need to create a teach-by-demonstration framework that captures the strengths of movement primitives and verifiable, safe control. We provide a framework that learns safe control laws from a probability distribution of robot trajectories through the use of advanced nonlinear control that incorporates safety constraints. Typically, such distributions are stochastic, making it difficult to offer any guarantees on safe operation. Our approach ensures that the distribution of allowed robot trajectories is within an envelope of safety and allows for robust operation of a robot. Furthermore, using our framework various probability distributions can be combined to represent complex scenarios in the environment. It will benefit practitioners by making it substantially easier to test and deploy accurate, efficient, and safe robots in complex real-world scenarios. The approach is currently limited to scenarios involving static obstacles, with dynamic obstacle avoidance an avenue of future effort.more » « less
-
Meta-learning owns unique effectiveness and swiftness in tackling emerging tasks with limited data. Its broad applicability is revealed by viewing it as a bi-level optimization problem. The resultant algorithmic viewpoint however, faces scalability issues when the inner-level optimization relies on gradient-based iterations. Implicit differentiation has been considered to alleviate this challenge, but it is restricted to an isotropic Gaussian prior, and only favors deterministic meta-learning approaches. This work markedly mitigates the scalability bottleneck by cross-fertilizing the benefits of implicit differentiation to probabilistic Bayesian meta-learning. The novel implicit Bayesian meta-learning (iBaML) method not only broadens the scope of learnable priors, but also quantifies the associated uncertainty. Furthermore, the ultimate complexity is well controlled regardless of the inner-level optimization trajectory. Analytical error bounds are established to demonstrate the precision and efficiency of the generalized implicit gradient over the explicit one. Extensive numerical tests are also carried out to empirically validate the performance of the proposed method.more » « less
-
Meta-learning owns unique effectiveness and swiftness in tackling emerging tasks with limited data. Its broad applicability is revealed by viewing it as a bi-level optimization problem. The resultant algorithmic viewpoint however, faces scalability issues when the inner-level optimization relies on gradient-based iterations. Implicit differentiation has been considered to alleviate this challenge, but it is restricted to an isotropic Gaussian prior, and only favors deterministic meta-learning approaches. This work markedly mitigates the scalability bottleneck by cross-fertilizing the benefits of implicit differentiation to probabilistic Bayesian meta-learning. The novel implicit Bayesian meta-learning (iBaML) method not only broadens the scope of learnable priors, but also quantifies the associated uncertainty. Furthermore, the ultimate complexity is well controlled regardless of the inner-level optimization trajectory. Analytical error bounds are established to demonstrate the precision and efficiency of the generalized implicit gradient over the explicit one. Extensive numerical tests are also carried out to empirically validate the performance of the proposed method.more » « less
-
Here, we report on the long-term stability of changes in behavior and brain activity following perceptual learning of conjunctions of simple motion features. Participants were trained for 3 weeks on a visual search task involving the detection of a dot moving in a “v”-shaped target trajectory among inverted “v”-shaped distractor trajectories. The first and last training sessions were carried out during functional magnetic resonance imaging (fMRI). Learning stability was again examined behaviorally and using fMRI 3 years after the end of training. Results show that acquired behavioral improvements were remarkably stable over time and that these changes were specific to trained target and distractor trajectories. A similar pattern was observed on the neuronal level, when the representation of target and distractor stimuli was examined in early retinotopic visual cortex (V1–V3): training enhanced activity for the target relative to the surrounding distractors in the search array and this enhancement persisted after 3 years. However, exchanging target and distractor trajectories abolished both neuronal and behavioral effects, suggesting that training-induced changes in stimulus representation are specific to trained stimulus identities.more » « less
