Think over Trajectories: Leveraging Video Generation to Reconstruct GPS Trajectories from Cellular Signaling

• URL: http://arxiv.org/abs/2603.26610v1
• Date: Fri, 27 Mar 2026 17:07:13 GMT
• Title: Think over Trajectories: Leveraging Video Generation to Reconstruct GPS Trajectories from Cellular Signaling
• Abstract summary: This paper studies the Sig2GPS problem: reconstructing GPS trajectories from cellular signaling.<n>To support this paradigm, a paired signaling-to-trajectory video dataset is constructed to fine-tune an open-source video model.<n>Experiments on large-scale real-world datasets show substantial improvements over strong engineered and learning-based baselines.
• Score: 24.925378866242426
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Mobile devices continuously interact with cellular base stations, generating massive volumes of signaling records that provide broad coverage for understanding human mobility. However, such records offer only coarse location cues (e.g., serving-cell identifiers) and therefore limit their direct use in applications that require high-precision GPS trajectories. This paper studies the Sig2GPS problem: reconstructing GPS trajectories from cellular signaling. Inspired by domain experts often lay the signaling trace on the map and sketch the corresponding GPS route, unlike conventional solutions that rely on complex multi-stage engineering pipelines or regress coordinates, Sig2GPS is reframed as an image-to-video generation task that directly operates in the map-visual domain: signaling traces are rendered on a map, and a video generation model is trained to draw a continuous GPS path. To support this paradigm, a paired signaling-to-trajectory video dataset is constructed to fine-tune an open-source video model, and a trajectory-aware reinforcement learning-based optimization method is introduced to improve generation fidelity via rewards. Experiments on large-scale real-world datasets show substantial improvements over strong engineered and learning-based baselines, while additional results on next GPS prediction indicate scalability and cross-city transferability. Overall, these results suggest that map-visual video generation provides a practical interface for trajectory data mining by enabling direct generation and refinement of continuous paths under map constraints.

Related papers

• Spotter: Efficient Urban Visual Localization via Geo-Referenced Facade Landmarks in GPS-Degraded Environments [0.5371337604556311]
  Spotter is a real-time visual localization framework for robotic and wearable platforms.<n>It processes Google Street View panoramas by semantically segmenting facades and pairing multi-view stereo depth with cartographic data.<n>At runtime, query images are matched via a cascaded retrieval and geometric verification pipeline to recover fine-grained global camera localization.
  arXiv  Detail & Related papers  (2026-08-24T14:15:25Z)
• CrossTracer: Cross-Embodiment Navigation via VLA Model Reasoning and Trace Residuals Adapting [51.52109090141233]
  CrossTracer is a hierarchical framework for cross-embodiment navigation.<n>It forms a unified pixel-space interface between semantic reasoning and physical grounding.<n>We evaluate CrossTracer on the NaviTrace benchmark, which tests whether a model can generate embodiment-consistent navigation traces.
  arXiv  Detail & Related papers  (2026-08-07T01:24:53Z)
• Geo-Align: Video Generation Alignment via Metric Geometry Reward [53.65904111864641]
  Geo-Align is the first Reinforcement Learning framework specifically designed for camera-controlled video re-rendering.<n>Built upon a pretrained model, we optimize the model through a scale-aware perceptual reward mechanism.<n>Experiments demonstrate that Geo-Align consistently outperforms existing supervised learning baselines in both precise camera controllability and visual fidelity.
  arXiv  Detail & Related papers  (2026-05-22T17:59:43Z)
• TrajFlow: Nation-wide Pseudo GPS Trajectory Generation with Flow Matching Models [18.786822812556625]
  TrajFlow is the first flow-matching-based generative model for GPS trajectory generation.<n>We show that TrajFlow consistently outperform diffusion-based and deep generative baselines at urban, metropolitan, and nationwide levels.<n>As the first nationwide, multi-scale GPS trajectory generation model, TrajFlow demonstrates strong potential to support inter-region urban planning, traffic management, and disaster response.
  arXiv  Detail & Related papers  (2026-03-16T09:15:13Z)
• Bayesian-Driven Graph Reasoning for Active Radio Map Construction [96.08082552413117]
  We propose an uncertainty-aware radio map reconstruction framework that explicitly leverages graph-based reasoning tailored for waypoint navigation.<n>Our approach integrates two key deep learning components: (1) a Bayesian neural network that estimates spatial uncertainty in real time, and (2) an attention-based reinforcement learning policy that performs global reasoning.<n> Experimental results show that URAM improves reconstruction accuracy by up to 34% over existing baselines.
  arXiv  Detail & Related papers  (2025-07-29T03:32:01Z)
• GPS as a Control Signal for Image Generation [95.43433150105385]
  We show that the GPS tags contained in photo metadata provide a useful control signal for image generation.<n>We train GPS-to-image models and use them for tasks that require a fine-grained understanding of how images vary within a city.
  arXiv  Detail & Related papers  (2025-01-21T18:59:46Z)
• GAReT: Cross-view Video Geolocalization with Adapters and Auto-Regressive Transformers [53.80009458891537]
  Cross-view video geo-localization aims to derive GPS trajectories from street-view videos by aligning them with aerial-view images. Current CVGL methods use camera and odometry data, typically absent in real-world scenarios. We propose GAReT, a fully transformer-based method for CVGL that does not require camera and odometry data.
  arXiv  Detail & Related papers  (2024-08-05T21:29:33Z)
• NLP-enabled Trajectory Map-matching in Urban Road Networks using a Transformer-based Encoder-decoder [1.3812010983144802]
  This study introduces a data-driven, deep learning-based map-matching framework, formulating the task as machine translation, inspired by NLP.<n>A transformer-based encoder-decoder model learns contextual representations of noisy GPS points to infer trajectory behavior and road structures in an end-to-end manner.<n>Experiments on synthetic trajectories show that this approach outperforms conventional methods by integrating contextual awareness.
  arXiv  Detail & Related papers  (2024-04-18T18:39:23Z)
• Leveraging Driver Field-of-View for Multimodal Ego-Trajectory Prediction [69.29802752614677]
  RouteFormer is a novel ego-trajectory prediction network combining GPS data, environmental context, and the driver's field-of-view.<n>To tackle data scarcity and enhance diversity, we introduce GEM, a dataset of urban driving scenarios enriched with synchronized driver field-of-view and gaze data.
  arXiv  Detail & Related papers  (2023-12-13T23:06:30Z)
• RNTrajRec: Road Network Enhanced Trajectory Recovery with Spatial-Temporal Transformer [15.350300338463969]
  We propose a road network enhanced transformer-based framework, namely RNTrajRec, for trajectory recovery. RNTrajRec first uses a graph model, namely GridGNN, to learn the embedding features of each road segment. It then introduces a Sub-Graph Generation module to represent each GPS point as a sub-graph structure of the road network around the GPS point.
  arXiv  Detail & Related papers  (2022-11-23T11:28:32Z)
• ViKiNG: Vision-Based Kilometer-Scale Navigation with Geographic Hints [94.60414567852536]
  Long-range navigation requires both planning and reasoning about local traversability. We propose a learning-based approach that integrates learning and planning. ViKiNG can leverage its image-based learned controller and goal-directed to navigate to goals up to 3 kilometers away.
  arXiv  Detail & Related papers  (2022-02-23T02:14:23Z)
• Multimodal Scale Consistency and Awareness for Monocular Self-Supervised Depth Estimation [1.1470070927586016]
  Self-supervised approaches on monocular videos suffer from scale-inconsistency across long sequences. We propose a dynamically-weighted GPS-to-Scale (g2s) loss to complement the appearance-based losses. We demonstrate scale-consistent and -aware depth estimation during inference, improving the performance even when training with low-frequency GPS data.
  arXiv  Detail & Related papers  (2021-03-03T15:39:41Z)

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.