MapTCL: Temporal Consistency Learning via Bidirectional Alignment for Vectorized HD Map Construction

• URL: http://arxiv.org/abs/2608.05209v1
• Date: Wed, 05 Aug 2026 08:53:26 GMT
• Title: MapTCL: Temporal Consistency Learning via Bidirectional Alignment for Vectorized HD Map Construction
• Abstract summary: We propose MapTCL, a training strategy that formulates temporal consistency loss between current and past frames.<n>By jointly training with these dual losses, MapTCL improves the temporal stability of generated HD maps.
• Score: 13.437125894280541
• License: http://creativecommons.org/publicdomain/zero/1.0/
• Abstract: Constructing reliable online HD maps remains challenging in dynamic urban environments due to moving objects and occlusions. While recent works employ feature-level temporal fusion to address this, they rely solely on per-frame ground truth supervision. Consequently, they lack an explicit objective to directly penalize the geometric noise and temporal jitter between consecutive online HD maps. To address this, we propose MapTCL, an auxiliary training strategy that formulates temporal consistency loss between current and past frames via bidirectional alignment. Specifically, Bidirectional Vector Consistency Learning (BVCL) models the geometric and semantic discrepancies between associated past and current vector instances as an auxiliary loss. We also employ Raster map Consistency Learning (RCL) as an additional loss to stabilize dense BEV features. By jointly training with these dual losses, MapTCL improves the temporal stability of generated HD maps. Extensive experiments on two standard benchmarks demonstrate the effectiveness of our approach. As a versatile plug-and-play module, MapTCL consistently enhances existing baseline models, achieving gains of +3.7 mAP & +2.8 C-mAP on nuScenes and +3.1 mAP & +2.5 C-mAP on Argoverse 2 without additional inference overhead.

Related papers

• OptiMVMap: Offline Vectorized Map Construction via Optimal Multi-vehicle Perspectives [111.69010751405368]
  offline vectorized maps constitute critical infrastructure for high-precision autonomous driving and mapping services.<n>Existing approaches rely predominantly on single ego-vehicle trajectories, which fundamentally suffer from viewpoint insufficiency.<n>We present OptiMVMap, which reformulates multi-vehicle mapping as a select-then-fuse problem to address these challenges systematically.
  arXiv  Detail & Related papers  (2026-04-18T20:23:27Z)
• MapGCLR: Geospatial Contrastive Learning of Representations for Online Vectorized HD Map Construction [44.37176678126708]
  Autonomous vehicles rely on map information to understand the world around them.<n>A more scalable alternative lies in online HD map construction, which only requires map annotations at training time.<n>This work focuses on improving the latent birds-eye-view (BEV) feature grid representation within a vectorized online HD map construction model.<n>We train the same model supervised using a reduced set of single-traversal labeled data and self-supervised on a broader unlabeled set of data following our multi-traversal requirements.
  arXiv  Detail & Related papers  (2026-03-11T11:58:23Z)
• DAMap: Distance-aware MapNet for High Quality HD Map Construction [64.36189046749038]
  Current methods perform poorly in high quality predictions due to inherent task misalignment.<n>We develop a novel HD map construction method named DAMap to address these problems.<n> DAMap consists of three components: Distance-aware Focal Loss (DAFL), Hybrid Loss Scheme (HLS), and Task Modulated Deformable Attention (TMDA)
  arXiv  Detail & Related papers  (2025-10-26T13:29:26Z)
• RTMap: Real-Time Recursive Mapping with Change Detection and Localization [8.343318095882232]
  RTMap persistently crowdsourcing a multi-traversal HD map as a self-evolutional memory.<n>On onboard agents, RTMap simultaneously addresses three core challenges in an end-to-end fashion.<n> Experiments on several public autonomous driving datasets demonstrate our solid performance on both the prior-aided map quality and the localization accuracy.
  arXiv  Detail & Related papers  (2025-07-01T17:32:30Z)
• TopoSD: Topology-Enhanced Lane Segment Perception with SDMap Prior [70.84644266024571]
  We propose to train a perception model to "see" standard definition maps (SDMaps) We encode SDMap elements into neural spatial map representations and instance tokens, and then incorporate such complementary features as prior information. Based on the lane segment representation framework, the model simultaneously predicts lanes, centrelines and their topology.
  arXiv  Detail & Related papers  (2024-11-22T06:13:42Z)
• GenMapping: Unleashing the Potential of Inverse Perspective Mapping for Robust Online HD Map Construction [20.1127163541618]
  We have designed a universal map generation framework, GenMapping. The framework is established with a triadic synergy architecture, including principal and dual auxiliary branches. A thorough array of experimental results shows that the proposed model surpasses current state-of-the-art methods in both semantic mapping and vectorized mapping, while also maintaining a rapid inference speed.
  arXiv  Detail & Related papers  (2024-09-13T10:15:28Z)
• DTCLMapper: Dual Temporal Consistent Learning for Vectorized HD Map Construction [20.6143278960295]
  This paper focuses on temporal instance consistency and temporal map consistency learning. DTCLMapper is a dual-stream temporal consistency learning module that combines instance embedding with geometry maps. Experiments on well-recognized benchmarks indicate that the proposed DTCLMapper achieves state-of-the-art performance in vectorized mapping tasks.
  arXiv  Detail & Related papers  (2024-05-09T02:58:55Z)
• ADMap: Anti-disturbance framework for reconstructing online vectorized HD map [9.218463154577616]
  This paper proposes the Anti-disturbance Map reconstruction framework (ADMap) To mitigate point-order jitter, the framework consists of three modules: Multi-Scale Perception Neck, Instance Interactive Attention (IIA), and Vector Direction Difference Loss (VDDL)
  arXiv  Detail & Related papers  (2024-01-24T01:37:27Z)
• Towards Real-World Visual Tracking with Temporal Contexts [64.7981374129495]
  We propose a two-level framework (TCTrack) that can exploit temporal contexts efficiently. Based on it, we propose a stronger version for real-world visual tracking, i.e., TCTrack++. For feature extraction, we propose an attention-based temporally adaptive convolution to enhance the spatial features. For similarity map refinement, we introduce an adaptive temporal transformer to encode the temporal knowledge efficiently.
  arXiv  Detail & Related papers  (2023-08-20T17:59:40Z)
• IDEA-Net: Dynamic 3D Point Cloud Interpolation via Deep Embedding Alignment [58.8330387551499]
  We formulate the problem as estimation of point-wise trajectories (i.e., smooth curves) We propose IDEA-Net, an end-to-end deep learning framework, which disentangles the problem under the assistance of the explicitly learned temporal consistency. We demonstrate the effectiveness of our method on various point cloud sequences and observe large improvement over state-of-the-art methods both quantitatively and visually.
  arXiv  Detail & Related papers  (2022-03-22T10:14:08Z)
• Robust Self-Supervised LiDAR Odometry via Representative Structure Discovery and 3D Inherent Error Modeling [67.75095378830694]
  We develop a two-stage odometry estimation network, where we obtain the ego-motion by estimating a set of sub-region transformations. In this paper, we aim to alleviate the influence of unreliable structures in training, inference and mapping phases. Our two-frame odometry outperforms the previous state of the arts by 16%/12% in terms of translational/rotational errors.
  arXiv  Detail & Related papers  (2022-02-27T12:52:27Z)

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.