SynthCity is a 367.9M point synthetic full-colour Mobile Laser Scanning point cloud with per-point labels from nine categories, generated in Blender for an urban environment.
SqueezeSegV2: Improved Model Structure and Unsupervised Domain Adaptation for Road-Object Segmentation from a LiDAR Point Cloud
3 Pith papers cite this work. Polarity classification is still indexing.
abstract
Earlier work demonstrates the promise of deep-learning-based approaches for point cloud segmentation; however, these approaches need to be improved to be practically useful. To this end, we introduce a new model SqueezeSegV2 that is more robust to dropout noise in LiDAR point clouds. With improved model structure, training loss, batch normalization and additional input channel, SqueezeSegV2 achieves significant accuracy improvement when trained on real data. Training models for point cloud segmentation requires large amounts of labeled point-cloud data, which is expensive to obtain. To sidestep the cost of collection and annotation, simulators such as GTA-V can be used to create unlimited amounts of labeled, synthetic data. However, due to domain shift, models trained on synthetic data often do not generalize well to the real world. We address this problem with a domain-adaptation training pipeline consisting of three major components: 1) learned intensity rendering, 2) geodesic correlation alignment, and 3) progressive domain calibration. When trained on real data, our new model exhibits segmentation accuracy improvements of 6.0-8.6% over the original SqueezeSeg. When training our new model on synthetic data using the proposed domain adaptation pipeline, we nearly double test accuracy on real-world data, from 29.0% to 57.4%. Our source code and synthetic dataset will be open-sourced.
fields
cs.CV 3years
2019 3verdicts
UNVERDICTED 3representative citing papers
A new CNN architecture for LiDAR semantic labeling achieves higher cross-sensor portability with a reported 10 percentage point IoU gain over a reference method.
A survey of deep learning architectures for 3D sensed data classification covering RGB-D, multi-view, volumetric and end-to-end methods along with datasets and future directions.
citing papers explorer
-
SynthCity: A large scale synthetic point cloud
SynthCity is a 367.9M point synthetic full-colour Mobile Laser Scanning point cloud with per-point labels from nine categories, generated in Blender for an urban environment.
-
Analyzing the Cross-Sensor Portability of Neural Network Architectures for LiDAR-based Semantic Labeling
A new CNN architecture for LiDAR semantic labeling achieves higher cross-sensor portability with a reported 10 percentage point IoU gain over a reference method.
-
A review on deep learning techniques for 3D sensed data classification
A survey of deep learning architectures for 3D sensed data classification covering RGB-D, multi-view, volumetric and end-to-end methods along with datasets and future directions.