Remembering Both the Machine and the Crowd When Sampling Points: Active Learning for Semantic Segmentation of ALS Point Clouds

Kölle, Michael; Walter, Volker; Schmohl, Stefan; Soergel, Uwe

doi:10.1007/978-3-030-68787-8_37

Cited by 7 publications

(14 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In case of our dense dataset, close neighbors of individual points incorporate very similar feature vectors. Since AL aims on selecting most informative points avoiding duplicates, our spatial subsampling not only boosts processing speed but also helps guaranteeing diversity of selected points (Kölle et al, 2021).…”

Section: Resultsmentioning

confidence: 99%

“…However, we need to keep in mind that non-experts (i.e., the crowd) are asked to generate labels for the selected points. We aim to evaluate whether we can ease interpretability of sampled points by further applying the method proposed in Kölle et al (2021), denoted as Reducing Interpretation Uncertainty (RIU). Sampled points are often situated exactly on class borders, where the true label is ambiguous and labeling is strongly dependent on the individual class understanding.…”

Section: Al Loop For Coupled Semantic Segmentation Of Pointmentioning

confidence: 99%

“…Sampled points are often situated exactly on class borders, where the true label is ambiguous and labeling is strongly dependent on the individual class understanding. Therefore, we try to increase the distance to the class border by considering the selected point as seed point and decide to use a point within distance dRIU instead when the sampling score (i.e., the weighted entropy value) of this instance is lower, assuming that a lower score is closely related to the distance to class boundary (detailed explanation can be found in Kölle et al (2021)). Afterwards the selected points (either applying RIU or not) are presented to crowdworkers for labeling.…”

Section: Al Loop For Coupled Semantic Segmentation Of Pointmentioning

confidence: 99%

“…Reference data for the point cloud was manually provided by the authors as outlined in Kölle et al (2021). For our study, we decide to only use a subset of available classes (i.e., we merge classes), since in it is observed that some classes are hard to interpret for non-experts.…”

Section: Creation Of Ground Truth Datamentioning

confidence: 99%

“…This means manual labeling is focused on most informative instances only. This approach of learning from a small pool of labeled points has already proven to be successful for semantic segmentation of point clouds, as demonstrated in Luo et al (2018), Li and Pfeifer (2019), Lin et al (2020) and Kölle et al (2021). While to the best of our knowledge for semantic interpretation of 3D meshes no AL approaches have been proposed in literature so far, some strategies for Pas-sive Learning (PL) are present.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Which 3d Data Representation Does the Crowd Like Best? Crowd-Based Active Learning for Coupled Semantic Segmentation of Point Clouds and Textured Meshes

Kölle

Laupheimer

Walter

et al. 2021

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Abstract. Semantic interpretation of multi-modal datasets is of great importance in many domains of geospatial data analysis. However, when training models for automated semantic segmentation, labeled training data is required and in case of multi-modality for each representation form of the scene. To completely avoid the time-consuming and cost-intensive involvement of an expert in the annotation procedure, we propose an Active Learning (AL) pipeline where a Random Forest classifier selects a subset of points sufficient for training and where necessary labels are received from the crowd. In this AL loop, we aim on coupled semantic segmentation of an Airborne Laser Scanning (ALS) point cloud and the corresponding 3D textured mesh generated from LiDAR data and imagery in a hybrid manner. Within this work we pursue two main objectives: i) We evaluate the performance of the AL pipeline applied to an ultra-high resolution ALS point cloud and a derived textured mesh (both benchmark datasets are available at https://ifpwww.ifp.uni-stuttgart.de/benchmark/hessigheim/default.aspx). ii) We investigate the capabilities of the crowd regarding interpretation of 3D geodata and observed that the crowd performs about 3 percentage points better when labeling meshes compared to point clouds. We additionally demonstrate that labels received solely by the crowd can power a machine learning system only differing in Overall Accuracy by less than 2 percentage points for the point cloud and less than 3 percentage points for the mesh, compared to using the completely labeled training pool. For deriving this sparse training set, we ask the crowd to label 0.25 % of available training points, resulting in costs of 190 &dollar;.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Al Loop For Coupled Semantic Segmentation Of Pointmentioning

confidence: 99%

Section: Al Loop For Coupled Semantic Segmentation Of Pointmentioning

confidence: 99%

Section: Creation Of Ground Truth Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Which 3d Data Representation Does the Crowd Like Best? Crowd-Based Active Learning for Coupled Semantic Segmentation of Point Clouds and Textured Meshes

Kölle

Laupheimer

Walter

et al. 2021

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

show abstract

CATEGORISE: An Automated Framework for Utilizing the Workforce of the Crowd for Semantic Segmentation of 3D Point Clouds

Kölle

Walter

Shiller

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Building a Fully-Automatized Active Learning Framework for the Semantic Segmentation of Geospatial 3D Point Clouds

Kölle,

Walter,

Sörgel

2024

PFG

View full text Add to dashboard Cite

In recent years, significant progress has been made in developing supervised Machine Learning (ML) systems like Convolutional Neural Networks. However, it’s crucial to recognize that the performance of these systems heavily relies on the quality of labeled training data. To address this, we propose a shift in focus towards developing sustainable methods of acquiring such data instead of solely building new classifiers in the ever-evolving ML field. Specifically, in the geospatial domain, the process of generating training data for ML systems has been largely neglected in research. Traditionally, experts have been burdened with the laborious task of labeling, which is not only time-consuming but also inefficient. In our system for the semantic interpretation of Airborne Laser Scanning point clouds, we break with this convention and completely remove labeling obligations from domain experts who have completed special training in geosciences and instead adopt a hybrid intelligence approach. This involves active and iterative collaboration between the ML model and humans through Active Learning, which identifies the most critical samples justifying manual inspection. Only these samples (typically $$\ll 1{\%}$$ ≪ 1 % of Passive Learning training points) are subject to human annotation. To carry out this annotation, we choose to outsource the task to a large group of non-specialists, referred to as the crowd, which comes with the inherent challenge of guiding those inexperienced annotators (i.e., “short-term employees”) to still produce labels of sufficient quality. However, we acknowledge that attracting enough volunteers for crowdsourcing campaigns can be challenging due to the tedious nature of labeling tasks. To address this, we propose employing paid crowdsourcing and providing monetary incentives to crowdworkers. This approach ensures access to a vast pool of prospective workers through respective platforms, ensuring timely completion of jobs. Effectively, crowdworkers become human processing units in our hybrid intelligence system mirroring the functionality of electronic processing units.

show abstract

Remembering Both the Machine and the Crowd When Sampling Points: Active Learning for Semantic Segmentation of ALS Point Clouds

Cited by 7 publications

References 21 publications

Which 3d Data Representation Does the Crowd Like Best? Crowd-Based Active Learning for Coupled Semantic Segmentation of Point Clouds and Textured Meshes

Which 3d Data Representation Does the Crowd Like Best? Crowd-Based Active Learning for Coupled Semantic Segmentation of Point Clouds and Textured Meshes

CATEGORISE: An Automated Framework for Utilizing the Workforce of the Crowd for Semantic Segmentation of 3D Point Clouds

Building a Fully-Automatized Active Learning Framework for the Semantic Segmentation of Geospatial 3D Point Clouds

Contact Info

Product

Resources

About