Contrast & Compress: Learning Lightweight Embeddings for Short Trajectories

• URL: http://arxiv.org/abs/2506.02571v1
• Date: Tue, 03 Jun 2025 07:53:04 GMT
• Title: Contrast & Compress: Learning Lightweight Embeddings for Short Trajectories
• Authors: Abhishek Vivekanandan, Christian Hubschneider, J. Marius Zöllner,
• Abstract summary: We propose a novel framework for learning fixed-dimensional embeddings for short trajectories by leveraging a Transformer encoder.<n>We analyze the influence of Cosine and FFT-based similarity metrics within the contrastive learning paradigm.<n>Our empirical evaluation on the Argoverse 2 dataset demonstrates that embeddings shaped by Cosine similarity objectives yield superior clustering of trajectories.
• Score: 11.6132604160666
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The ability to retrieve semantically and directionally similar short-range trajectories with both accuracy and efficiency is foundational for downstream applications such as motion forecasting and autonomous navigation. However, prevailing approaches often depend on computationally intensive heuristics or latent anchor representations that lack interpretability and controllability. In this work, we propose a novel framework for learning fixed-dimensional embeddings for short trajectories by leveraging a Transformer encoder trained with a contrastive triplet loss that emphasize the importance of discriminative feature spaces for trajectory data. We analyze the influence of Cosine and FFT-based similarity metrics within the contrastive learning paradigm, with a focus on capturing the nuanced directional intent that characterizes short-term maneuvers. Our empirical evaluation on the Argoverse 2 dataset demonstrates that embeddings shaped by Cosine similarity objectives yield superior clustering of trajectories by both semantic and directional attributes, outperforming FFT-based baselines in retrieval tasks. Notably, we show that compact Transformer architectures, even with low-dimensional embeddings (e.g., 16 dimensions, but qualitatively down to 4), achieve a compelling balance between retrieval performance (minADE, minFDE) and computational overhead, aligning with the growing demand for scalable and interpretable motion priors in real-time systems. The resulting embeddings provide a compact, semantically meaningful, and efficient representation of trajectory data, offering a robust alternative to heuristic similarity measures and paving the way for more transparent and controllable motion forecasting pipelines.

Related papers

• TraF-Align: Trajectory-aware Feature Alignment for Asynchronous Multi-agent Perception [7.382491303268417]
  TraF-Align learns the flow path of features by predicting the feature-level trajectory of objects from past observations up to the ego vehicle's current time.<n>This approach corrects spatial misalignment and ensures semantic consistency across agents, effectively compensating for motion.<n>Experiments on two real-world datasets, V2V4Real and DAIR-V2X-Seq, show that TraF-Align sets a new benchmark for asynchronous cooperative perception.
  arXiv  Detail & Related papers  (2025-03-25T06:56:35Z)
• T-JEPA: A Joint-Embedding Predictive Architecture for Trajectory Similarity Computation [6.844357745770191]
  Trajectory similarity computation is an essential technique for analyzing moving patterns of spatial data across various applications. We propose T-JEPA, a self-supervised trajectory similarity method employing Joint-Embedding Predictive Architecture (JEPA) to enhance trajectory representation learning.
  arXiv  Detail & Related papers  (2024-06-13T09:51:51Z)
• Hallmarks of Optimization Trajectories in Neural Networks: Directional Exploration and Redundancy [75.15685966213832]
  We analyze the rich directional structure of optimization trajectories represented by their pointwise parameters. We show that training only scalar batchnorm parameters some while into training matches the performance of training the entire network.
  arXiv  Detail & Related papers  (2024-03-12T07:32:47Z)
• Disentangled Representation Learning with Transmitted Information Bottleneck [57.22757813140418]
  We present textbfDisTIB (textbfTransmitted textbfInformation textbfBottleneck for textbfDisd representation learning), a novel objective that navigates the balance between information compression and preservation.
  arXiv  Detail & Related papers  (2023-11-03T03:18:40Z)
• Intensity Profile Projection: A Framework for Continuous-Time Representation Learning for Dynamic Networks [50.2033914945157]
  We present a representation learning framework, Intensity Profile Projection, for continuous-time dynamic network data. The framework consists of three stages: estimating pairwise intensity functions, learning a projection which minimises a notion of intensity reconstruction error. Moreoever, we develop estimation theory providing tight control on the error of any estimated trajectory, indicating that the representations could even be used in quite noise-sensitive follow-on analyses.
  arXiv  Detail & Related papers  (2023-06-09T15:38:25Z)
• Motion-Scenario Decoupling for Rat-Aware Video Position Prediction: Strategy and Benchmark [49.58762201363483]
  We introduce RatPose, a bio-robot motion prediction dataset constructed by considering the influence factors of individuals and environments. We propose a Dual-stream Motion-Scenario Decoupling framework that effectively separates scenario-oriented and motion-oriented features. We demonstrate significant performance improvements of the proposed textitDMSD framework on different difficulty-level tasks.
  arXiv  Detail & Related papers  (2023-05-17T14:14:31Z)
• Contrastive Trajectory Similarity Learning with Dual-Feature Attention [24.445998309807965]
  Tray similarity measures act as query predicates in trajectory databases. We propose a contrastive learning-based trajectory modelling method named TrajCL. TrajCL is consistently and significantly more accurate and faster than the state-of-the-art trajectory similarity measures.
  arXiv  Detail & Related papers  (2022-10-11T05:25:14Z)
• Imposing Consistency for Optical Flow Estimation [73.53204596544472]
  Imposing consistency through proxy tasks has been shown to enhance data-driven learning. This paper introduces novel and effective consistency strategies for optical flow estimation.
  arXiv  Detail & Related papers  (2022-04-14T22:58:30Z)
• ST2Vec: Spatio-Temporal Trajectory Similarity Learning in Road Networks [27.452831603278565]
  We propose ST2Vec, a trajectory-learning based architecture that considers fine-grained spatial and temporal between pairs of trajectories. Inspired by curriculum concept, ST2Vec employs curriculum learning for model optimization to improve both convergence and effectiveness. An experimental study offers evidence that ST2Vec outperforms all state-of-the-art competitors substantially in terms of effectiveness, efficiency, and robustness.
  arXiv  Detail & Related papers  (2021-12-17T06:18:04Z)
• SGCN:Sparse Graph Convolution Network for Pedestrian Trajectory Prediction [64.16212996247943]
  We present a Sparse Graph Convolution Network(SGCN) for pedestrian trajectory prediction. Specifically, the SGCN explicitly models the sparse directed interaction with a sparse directed spatial graph to capture adaptive interaction pedestrians. visualizations indicate that our method can capture adaptive interactions between pedestrians and their effective motion tendencies.
  arXiv  Detail & Related papers  (2021-04-04T03:17:42Z)
• An Unsupervised Learning Method with Convolutional Auto-Encoder for Vessel Trajectory Similarity Computation [13.003061329076775]
  We propose an unsupervised learning method which automatically extracts low-dimensional features through a convolutional auto-encoder (CAE) Based on the massive vessel trajectories collected, the CAE can learn the low-dimensional representations of informative trajectory images in an unsupervised manner. The proposed method largely outperforms traditional trajectory similarity methods in terms of efficiency and effectiveness.
  arXiv  Detail & Related papers  (2021-01-10T04:42:11Z)

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.