TSD: A Physics-Inspired Trajectory Saliency Detector for Efficient Imitation Learning

• URL: http://arxiv.org/abs/2606.23371v1
• Date: Mon, 22 Jun 2026 14:06:04 GMT
• Title: TSD: A Physics-Inspired Trajectory Saliency Detector for Efficient Imitation Learning
• Abstract summary: Trajectory Saliency Detector (TSD) is a training-free and plug-and-play framework to identify trajectory saliency.<n>TSD employs two physically-grounded metrics: spatial entropy to capture fine-grained manipulation and centripetal acceleration to detect agile maneuvering.<n>Extensive experiments in both simulation and real-world settings demonstrate that models trained on TSD-condensed datasets achieve comparable or even superior performance with 25% less data on average.
• Score: 9.737683365003322
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: For imitation learning in robotic manipulation, high data collection costs result in the scarcity of high quality data. In this paper, we leverage the inherent heterogeneity of trajectories to address this challenge. Based on our observations of manipulation tasks, we categorize motions into transitional, precise, and agile types, defining the latter two as trajectory saliency due to their criticality to task success in contrast to the prevalent but less relevant transitional motions. Therefore, we propose the Trajectory Saliency Detector (TSD), a training-free and plug-and-play framework to identify trajectory saliency. TSD employs two physically-grounded metrics: spatial entropy to capture fine-grained manipulation and centripetal acceleration to detect agile maneuvering. We further leverage TSD to develop a dataset compression method that reduces training costs and a dataset expansion strategy that improves data collection efficiency. Extensive experiments in both simulation and real-world settings demonstrate that models trained on TSD-condensed datasets achieve comparable or even superior performance with 25% less data on average. These results validate the effectiveness of our dataset compression and expansion strategies, thereby confirming the utility of TSD. Consequently, TSD offers a scalable and cost-effective pathway to synthesize information-dense datasets for efficient robot learning. Project page: https://trajectory-saliency-detector.github.io/trajectory-saliency-detector/

Related papers

• SC-GIR: Goal-oriented Semantic Communication via Invariant Representation Learning [59.45312293893698]
  Goal-oriented semantic communication (SC) aims to revolutionize communication systems by transmitting only task-essential information.<n>We propose a novel framework called Goal-oriented Invariant Representation-based SC (SC-GIR) for image transmission.
  arXiv  Detail & Related papers  (2025-09-01T04:29:43Z)
• Latent Factorization of Tensors with Threshold Distance Weighted Loss for Traffic Data Estimation [4.079031335530995]
  In real-word traffic data collection processes, issues such as communication failures often lead to incomplete or corrupted datasets.<n>Latent factorization of outliers (LFT) model has emerged as widely adopted and effective solution.<n>This paper proposes a threshold distance weighted (TDW) loss sensitivity-ind Latent Factorization of outliers (TDFTWL) model.<n>The proposed TDFTWL model consistently outperforms state-of-the-art approaches in terms of both accuracy and computational efficiency.
  arXiv  Detail & Related papers  (2025-06-11T05:36:13Z)
• SUDS: A Strategy for Unsupervised Drift Sampling [0.5437605013181142]
  Supervised machine learning encounters concept drift, where the data distribution changes over time, degrading performance. We present the Strategy for Drift Sampling (SUDS), a novel method that selects homogeneous samples for retraining using existing drift detection algorithms. Our results demonstrate the efficacy of SUDS in optimizing labeled data use in dynamic environments.
  arXiv  Detail & Related papers  (2024-11-05T10:55:29Z)
• Towards Stable and Storage-efficient Dataset Distillation: Matching Convexified Trajectory [53.37473225728298]
  The rapid evolution of deep learning and large language models has led to an exponential growth in the demand for training data. Matching Training Trajectories (MTT) has been a prominent approach, which replicates the training trajectory of an expert network on real data with a synthetic dataset. We introduce a novel method called Matching Convexified Trajectory (MCT), which aims to provide better guidance for the student trajectory.
  arXiv  Detail & Related papers  (2024-06-28T11:06:46Z)
• Energy-Efficient Edge Learning via Joint Data Deepening-and-Prefetching [9.468399367975984]
  We propose a novel offloading architecture called joint data deepening-and-prefetching (JD2P) JD2P is feature-by-feature offloading comprising two key techniques. We evaluate the effectiveness of JD2P through experiments using the MNIST dataset.
  arXiv  Detail & Related papers  (2024-02-19T08:12:47Z)
• ALF: Adaptive Label Finetuning for Scene Graph Generation [116.59868289196157]
  Scene Graph Generation endeavors to predict the relationships between subjects and objects in a given image. Long-tail distribution of relations often leads to biased prediction on coarse labels, presenting a substantial hurdle in SGG. We introduce one-stage data transfer pipeline in SGG, termed Adaptive Label Finetuning (ALF), which eliminates the need for extra retraining sessions. ALF achieves a 16% improvement in mR@100 compared to the typical SGG method Motif, with only a 6% increase in calculation costs compared to the state-of-the-art method IETrans.
  arXiv  Detail & Related papers  (2023-12-29T01:37:27Z)
• Enhancing IoT Security via Automatic Network Traffic Analysis: The Transition from Machine Learning to Deep Learning [0.0]
  This work provides a comparative analysis illustrating how Deep Learning (DL) surpasses Machine Learning (ML) in addressing tasks within Internet of Things (IoT) Our approach involves training and evaluating a DL model using a range of diverse IoT-related datasets. Experiments showcase the ability of DL to surpass the constraints tied to manually engineered features, achieving superior results in attack detection and maintaining comparable outcomes in device-type identification.
  arXiv  Detail & Related papers  (2023-11-20T16:48:50Z)
• AST: Effective Dataset Distillation through Alignment with Smooth and High-Quality Expert Trajectories [18.266786462036553]
  We propose an effective DD framework named AST, standing for Alignment with Smooth and high-quality expert Trajectories. We conduct extensive experiments on datasets of different scales, sizes, and resolutions.
  arXiv  Detail & Related papers  (2023-10-16T16:13:53Z)
• Leveraging the Power of Data Augmentation for Transformer-based Tracking [64.46371987827312]
  We propose two data augmentation methods customized for tracking. First, we optimize existing random cropping via a dynamic search radius mechanism and simulation for boundary samples. Second, we propose a token-level feature mixing augmentation strategy, which enables the model against challenges like background interference.
  arXiv  Detail & Related papers  (2023-09-15T09:18:54Z)
• Fast Machine Unlearning Without Retraining Through Selective Synaptic Dampening [51.34904967046097]
  Selective Synaptic Dampening (SSD) is a fast, performant, and does not require long-term storage of the training data. We present a novel two-step, post hoc, retrain-free approach to machine unlearning which is fast, performant, and does not require long-term storage of the training data.
  arXiv  Detail & Related papers  (2023-08-15T11:30:45Z)
• Minimizing the Accumulated Trajectory Error to Improve Dataset Distillation [151.70234052015948]
  We propose a novel approach that encourages the optimization algorithm to seek a flat trajectory. We show that the weights trained on synthetic data are robust against the accumulated errors perturbations with the regularization towards the flat trajectory. Our method, called Flat Trajectory Distillation (FTD), is shown to boost the performance of gradient-matching methods by up to 4.7%.
  arXiv  Detail & Related papers  (2022-11-20T15:49:11Z)
• PreTraM: Self-Supervised Pre-training via Connecting Trajectory and Map [58.53373202647576]
  We propose PreTraM, a self-supervised pre-training scheme for trajectory forecasting. It consists of two parts: 1) Trajectory-Map Contrastive Learning, where we project trajectories and maps to a shared embedding space with cross-modal contrastive learning, and 2) Map Contrastive Learning, where we enhance map representation with contrastive learning on large quantities of HD-maps. On top of popular baselines such as AgentFormer and Trajectron++, PreTraM boosts their performance by 5.5% and 6.9% relatively in FDE-10 on the challenging nuScenes dataset.
  arXiv  Detail & Related papers  (2022-04-21T23:01:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.