Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Co-PLNet: A Collaborative Point-Line Network for Prompt-Guided Wireframe Parsing

Chao Wang, Xuanying Li, Cheng Dai, Jinglei Feng, Yuxiang Luo, Yuqi Ouyang, Hao Qin

TL;DR

The paper tackles wireframe parsing, where accurate and consistent detection of line segments and junctions is essential for downstream SLAM. It introduces Co-PLNet, a collaborative framework that uses a Point-Line Prompt Encoder to generate spatial prompts from preliminary predictions and a Cross-Guidance Line Decoder that refines both lines and junctions via sparse attention. The main contributions are the point-line collaborative paradigm, prompt-based fusion of geometric cues, and an efficient attention mechanism that preserves real-time performance. Experiments on the Wireframe and YorkUrban datasets demonstrate competitive accuracy and robustness, achieving up to 76.8 FPS and showing that mutual prompts reduce endpoint mismatches and improve geometry consistency.

Abstract

Wireframe parsing aims to recover line segments and their junctions to form a structured geometric representation useful for downstream tasks such as Simultaneous Localization and Mapping (SLAM). Existing methods predict lines and junctions separately and reconcile them post-hoc, causing mismatches and reduced robustness. We present Co-PLNet, a point-line collaborative framework that exchanges spatial cues between the two tasks, where early detections are converted into spatial prompts via a Point-Line Prompt Encoder (PLP-Encoder), which encodes geometric attributes into compact and spatially aligned maps. A Cross-Guidance Line Decoder (CGL-Decoder) then refines predictions with sparse attention conditioned on complementary prompts, enforcing point-line consistency and efficiency. Experiments on Wireframe and YorkUrban show consistent improvements in accuracy and robustness, together with favorable real-time efficiency, demonstrating our effectiveness for structured geometry perception.

Co-PLNet: A Collaborative Point-Line Network for Prompt-Guided Wireframe Parsing

TL;DR

The paper tackles wireframe parsing, where accurate and consistent detection of line segments and junctions is essential for downstream SLAM. It introduces Co-PLNet, a collaborative framework that uses a Point-Line Prompt Encoder to generate spatial prompts from preliminary predictions and a Cross-Guidance Line Decoder that refines both lines and junctions via sparse attention. The main contributions are the point-line collaborative paradigm, prompt-based fusion of geometric cues, and an efficient attention mechanism that preserves real-time performance. Experiments on the Wireframe and YorkUrban datasets demonstrate competitive accuracy and robustness, achieving up to 76.8 FPS and showing that mutual prompts reduce endpoint mismatches and improve geometry consistency.

Abstract

Wireframe parsing aims to recover line segments and their junctions to form a structured geometric representation useful for downstream tasks such as Simultaneous Localization and Mapping (SLAM). Existing methods predict lines and junctions separately and reconcile them post-hoc, causing mismatches and reduced robustness. We present Co-PLNet, a point-line collaborative framework that exchanges spatial cues between the two tasks, where early detections are converted into spatial prompts via a Point-Line Prompt Encoder (PLP-Encoder), which encodes geometric attributes into compact and spatially aligned maps. A Cross-Guidance Line Decoder (CGL-Decoder) then refines predictions with sparse attention conditioned on complementary prompts, enforcing point-line consistency and efficiency. Experiments on Wireframe and YorkUrban show consistent improvements in accuracy and robustness, together with favorable real-time efficiency, demonstrating our effectiveness for structured geometry perception.
Paper Structure (12 sections, 11 equations, 3 figures, 3 tables)

This paper contains 12 sections, 11 equations, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Proposed Method
  3. Point-Line Collaborative Paradigm
  4. Point-Line Prompt Encoder
  5. Cross-Guidance Line Decoder
  6. Experiments
  7. Datasets and Evaluation Metrics
  8. Implementation Details
  9. Comparison with SOTA Methods
  10. Ablation Study
  11. Parameter Study
  12. Conclusion

Figures (3)

  • Figure 1: Conceptual comparison between existing wireframe parsing paradigms and our Co-PLNet. Note that this figure emphasizes architectural differences rather than visual output quality
  • Figure 2: Overview of the Co-PLNet framework. The PLP-Encoder generates spatial prompts from junction and line predictions, which are refined by the CGL-Decoder to produce accurate line segment predictions.
  • Figure 3: Visualization of parsing results. The first and second rows show images from the Wireframe and YorkUrban datasets, respectively, with predicted wireframes aligned to ground-truth annotations.