Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 24, 2025IEEE Transactions on Pattern Analysis and Machine Intelligence

Translating Images to Road Network: A Sequence-to-Sequence Perspective

View Full Paper
Ask AI
Bookmark
Share

Authors

JLJiachen LuMNMing NieBZBozhou Zhang

Discussion

Loading...

Member takes

Overview

This study proposes a non-autoregressive sequence-to-sequence model to improve road network extraction and landmark detection.

Key Points

  • The proposed non-autoregressive approach enhances the efficiency and accuracy of road network extraction.
  • RoadNet Sequence effectively merges Euclidean and non-Euclidean data for improved topology reasoning.
  • Experiments on the nuScenes dataset demonstrate superior performance compared to state-of-the-art methods.
  • Topology-Inherited Training addresses bottlenecks in landmark detection and topology reasoning.

Cite This Study

Lu et al. (2025) studied this question.

synapsesocial.com/papers/68d6d82e8b2b6861e4c3e2eahttps://doi.org/10.1109/tpami.2025.3612940
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Translating Images to Road Network:A Non-Autoregressive Sequence-to-Sequence Approach2024
  2. 2TransRoadNet: a Transformer framework for multi-modal road network pattern recognition2025
  3. 3TransRoadNet: a Transformer framework for multi-modal road network pattern recognition2025
  4. 4RoadFormer: Duplex Transformer for RGB-Normal Semantic Road Scene Parsing2024 · 58 citations
  5. 5A novel transformer‐based graph generation model for vectorized road design2024 · 1 citations