SIGMA-Lane: Scale-pyramId Gated MAmba for Temporally Consistent Video Lane Detection
2026-08-17 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors address the problem of detecting road lanes in videos when parts of the lanes are hidden by other vehicles, which can confuse the system over time. They propose SIGMA-Lane, a method that carefully controls how new, possibly corrupted information updates the model's memory to avoid mistakes lingering in future predictions. Their approach also uses past lane information aligned to the current view to help fill in missing lane details. Tests show SIGMA-Lane improves the consistency of lane detection even when heavy occlusions occur.
video lane detectiontemporal stabilityvehicle occlusionState Space Model (SSM)state contaminationgating mechanismstemporal memoryStructural Spatial Retrieval (SSR)affine alignmentF1 score
Authors
Tiancheng Zhang, Mengmeng Wang, Yan Gao, Xiangjie Kong, Guojiang Shen, Jiaxin Du
Abstract
Video lane detection requires predictions that remain stable across frames, yet severe vehicle occlusions can break temporal cues. In streaming recurrent models, corrupted observations may enter the hidden state and produce errors that persist into later frames. Existing occlusion-aware refinements usually provide obstacle masks as auxiliary inputs, so the state-update path is only indirectly protected. We propose SIGMA-Lane, which treats this failure mode as state contamination in State Space Model (SSM)-based temporal modeling. SIGMA-Lane places occlusion-aware gates on the SSM write and residual-fusion paths, controlling how current observations enter temporal memory and are fused back after temporal propagation. After coordinate-consistent affine alignment, the model combines two complementary paths: SSM-consistent dual-gating for temporal filtering and Structural Spatial Retrieval (SSR) for recovering missing lane structure from aligned historical priors. Experiments on VIL-100 and OpenLane-V show improved temporal stability under heavy occlusion, with competitive F1 and mIoU scores.