Wireframe Parsing With Guidance of Distance Map

Huang, Kun; Gao, Shenghua

doi:10.1109/access.2019.2943885

Cited by 6 publications

(6 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our fastest model, M-LSD-tiny with 320 input size, has a slightly lower performance than that of TP-LSD-Lite, but achieves an increase of 130.5% in inference speed with only 2.5% of the model size. Compared to the previous lightest model TP-LSD-HG, M-LSD with 512 input size outperforms on sAP 5 , sAP 10 and LAP with an increase of 136.0% in inference speed with 20.3% of the model size. Our lightest model, M-LSD-tiny with 320 input size, shows an increase of 310.6% in the inference speed with 8.1% of the model size compared to TP-LSD-HG.…”

Section: Comparison With Other Methodsmentioning

confidence: 84%

“…The top-down strategy [30] first detects regions of line segment with attraction field maps and then predicts line segments by squeezing regions into line segments. In contrast, the bottomup strategy first detects junctions, then arranges them into line segments, and lastly verifies the line segments by using an extra classifier [36,31,35] or a merging algorithm [10,11]. Recently, [12] proposes Tri-Points (TP) representation for a simpler process of line prediction without the time-consuming steps of line proposal and verification.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards Light-weight and Real-time Line Segment Detection

Gu¹,

Ko²,

Go³

et al. 2021

Preprint

View full text Add to dashboard Cite

Previous deep learning-based line segment detection (LSD) suffer from the immense model size and high computational cost for line prediction. This constrains them from real-time inference on computationally restricted environments. In this paper, we propose a real-time and lightweight line segment detector for resource-constrained environments named Mobile LSD (M-LSD). We design an extremely efficient LSD architecture by minimizing the backbone network and removing the typical multi-module process for line prediction in previous methods. To maintain competitive performance with such a light-weight network, we present novel training schemes: Segments of Line segment (SoL) augmentation and geometric learning scheme. SoL augmentation splits a line segment into multiple subparts, which are used to provide auxiliary line data during the training process. Moreover, the geometric learning scheme allows a model to capture additional geometric cues from matching loss, junction and line segmentation, length and degree regression. Compared with TP-LSD-Lite, previously the best real-time LSD method, our model (M-LSDtiny) achieves competitive performance with 2.5% of model size and an increase of 130.5% in inference speed on GPU when evaluated with Wireframe and YorkUrban datasets. Furthermore, our model runs at 56.8 FPS and 48.6 FPS on Android and iPhone mobile devices, respectively. To the best of our knowledge, this is the first real-time deep LSD method available on mobile devices.

show abstract

Section: Comparison With Other Methodsmentioning

confidence: 84%

Section: Introductionmentioning

confidence: 99%

Towards Light-weight and Real-time Line Segment Detection

Gu¹,

Ko²,

Go³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…From the user perspective, the ultimate goal would be to reverseengineer a CAD model (Gonzalez-Aguilera et al, 2012;Li et al, 2017b;Durupt et al, 2008). Several works have investigated wireframe parsing in 2D images (Huang et al, 2018;Huang & Gao, 2019). LCNN (Zhou et al, 2019a) is an end-to-end trainable system that directly outputs a vectorised wireframe.…”

Section: Related Workmentioning

confidence: 99%

PC2WF: 3D Wireframe Reconstruction from Raw Point Clouds

Liu¹,

D’Aronco²,

Schindler³

et al. 2021

Preprint

View full text Add to dashboard Cite

We introduce PC2WF, the first end-to-end trainable deep network architecture to convert a 3D point cloud into a wireframe model. The network takes as input an unordered set of 3D points sampled from the surface of some object, and outputs a wireframe of that object, i.e., a sparse set of corner points linked by line segments. Recovering the wireframe is a challenging task, where the numbers of both vertices and edges are different for every instance, and a-priori unknown. Our architecture gradually builds up the model: It starts by encoding the points into feature vectors. Based on those features, it identifies a pool of candidate vertices, then prunes those candidates to a final set of corner vertices and refines their locations. Next, the corners are linked with an exhaustive set of candidate edges, which is again pruned to obtain the final wireframe. All steps are trainable, and errors can be backpropagated through the entire sequence. We validate the proposed model on a publicly available synthetic dataset, for which the ground truth wireframes are accessible, as well as on a new real-world dataset. Our model produces wireframe abstractions of good quality and outperforms several baselines.

show abstract

“…The top-down strategy (Xue et al 2019a) first detects regions of line segment with attraction field maps and then squeezes these regions into line segments to make predictions. In contrast, the bottom-up strategy first detects junctions, then arranges them into line segments, and lastly verifies the line segments by using an extra classifier (Zhou, Qi, and Ma 2019;Xue et al 2020;Zhang et al 2019) or a merging algorithm (Huang and Gao 2019;Huang et al 2018). Recently, (Huang et al 2020) proposes Tri-Points (TP) representation for a simpler process of line prediction without the time-consuming steps of line proposal and verification.…”

Section: Introductionmentioning

confidence: 99%

Towards Light-Weight and Real-Time Line Segment Detection

et al. 2022

AAAI

View full text Add to dashboard Cite

Previous deep learning-based line segment detection (LSD) suffers from the immense model size and high computational cost for line prediction. This constrains them from real-time inference on computationally restricted environments. In this paper, we propose a real-time and light-weight line segment detector for resource-constrained environments named Mobile LSD (M-LSD). We design an extremely efficient LSD architecture by minimizing the backbone network and removing the typical multi-module process for line prediction found in previous methods. To maintain competitive performance with a light-weight network, we present novel training schemes: Segments of Line segment (SoL) augmentation, matching and geometric loss. SoL augmentation splits a line segment into multiple subparts, which are used to provide auxiliary line data during the training process. Moreover, the matching and geometric loss allow a model to capture additional geometric cues. Compared with TP-LSD-Lite, previously the best real-time LSD method, our model (M-LSD-tiny) achieves competitive performance with 2.5% of model size and an increase of 130.5% in inference speed on GPU. Furthermore, our model runs at 56.8 FPS and 48.6 FPS on the latest Android and iPhone mobile devices, respectively. To the best of our knowledge, this is the first real-time deep LSD available on mobile devices.

show abstract

Wireframe Parsing With Guidance of Distance Map

Cited by 6 publications

References 48 publications

Towards Light-weight and Real-time Line Segment Detection

Towards Light-weight and Real-time Line Segment Detection

PC2WF: 3D Wireframe Reconstruction from Raw Point Clouds

Towards Light-Weight and Real-Time Line Segment Detection

Contact Info

Product

Resources

About