PredMapNet: Future and Historical Reasoning for Consistent Online HD Vectorized Map Construction
2026-02-18 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors created a new system to build high-definition maps that help self-driving cars understand their surroundings more reliably over time. Their method starts by using smart queries that know about the scene's layout and remembers detailed past information about each map part. They also predict where things on the map will move shortly to keep everything consistent and avoid mistakes. Tests on popular datasets showed their approach works better and faster than previous methods.
HD mapsautonomous drivingquery-based methodssemantic masksmap trackingtemporal consistencyrasterized mapshort-term predictionnuScenesArgoverse2
Authors
Bo Lang, Nirav Savaliya, Zhihao Zheng, Jinglun Feng, Zheng-Hang Yeh, Mooi Choo Chuah
Abstract
High-definition (HD) maps are crucial to autonomous driving, providing structured representations of road elements to support navigation and planning. However, existing query-based methods often employ random query initialization and depend on implicit temporal modeling, which lead to temporal inconsistencies and instabilities during the construction of a global map. To overcome these challenges, we introduce a novel end-to-end framework for consistent online HD vectorized map construction, which jointly performs map instance tracking and short-term prediction. First, we propose a Semantic-Aware Query Generator that initializes queries with spatially aligned semantic masks to capture scene-level context globally. Next, we design a History Rasterized Map Memory to store fine-grained instance-level maps for each tracked instance, enabling explicit historical priors. A History-Map Guidance Module then integrates rasterized map information into track queries, improving temporal continuity. Finally, we propose a Short-Term Future Guidance module to forecast the immediate motion of map instances based on the stored history trajectories. These predicted future locations serve as hints for tracked instances to further avoid implausible predictions and keep temporal consistency. Extensive experiments on the nuScenes and Argoverse2 datasets demonstrate that our proposed method outperforms state-of-the-art (SOTA) methods with good efficiency.