LiDAR few-shot domain adaptation via integrated CycleGAN and 3D object detector with joint learning delay

Corral-Soto, Eduardo R.; Nabatchian, Amir; Gerdzhev, Martin; Liu, Bingbing

doi:10.1109/icra48506.2021.9561466

Cited by 7 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These splits are chosen deterministically using dataset sampling parameters available in the OpenPCDet [29] base code that we use in this paper. To avoid introducing domain shifts [18], [19], [20] between the dataset splits due to resampling of the point clouds at the input, we apply the resampling R(P, ⃗ θ) to both, the training and validation splits. We train and evaluate each detector on all of the following object classes: Waymo: {vehicle, pedestrian, cyclist}, ONCE: {car, bus, truck, pedestrian, cyclist}.…”

Section: A Experiments With Different 3d Object Detectorsmentioning

confidence: 99%

“…1: mAP for Centerpoint 3D object detector [7] trained/evaluated on Waymo point clouds with and without random sampling with 50% keep percentage on distance range [0-10] m. We computed mAP separately for different distance ranges, and observed that performace improved at farther away ranges, which suggests the existence of a learning bias toward the closer denser objects. Domain adaptation methods [18], [19], [20] have been used as a form of data augmentation by shifting the distribution of a large labeled source training dataset into that of a small target dataset to increase the amount of training data. Recent GAN-based object-level shape completion methods [21], [22], [23] have also been used to either complete occluded point clouds, or manipulate their distribution to attempt improving performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Domain Adaptation in LiDAR Semantic Segmentation via Hybrid Learning with Alternating Skip Connections

Corral-Soto

Rochan

et al. 2023

2023 IEEE Intelligent Vehicles Symposium (IV)

Self Cite

View full text Add to dashboard Cite

In recent years, much progress has been made in LiDAR-based 3D object detection mainly due to advances in detector architecture designs and availability of large-scale LiDAR datasets. Existing 3D object detectors tend to perform well on the point cloud regions closer to the LiDAR sensor as opposed to on regions that are farther away. In this paper, we investigate this problem from the data perspective instead of detector architecture design. We observe that there is a learning bias in detection models towards the dense objects near the sensor and show that the detection performance can be improved by simply manipulating the input point cloud density at different distance ranges without modifying the detector architecture and without data augmentation. We propose a modelfree point cloud density adjustment pre-processing mechanism that uses iterative MCMC optimization to estimate optimal parameters for altering the point density at different distance ranges. We conduct experiments using four state-of-the-art LiDAR 3D object detectors on two public LiDAR datasets, namely Waymo and ONCE. Our results demonstrate that our range-based point cloud density manipulation technique can improve the performance of the existing detectors, which in turn could potentially inspire future detector designs.

show abstract

Section: A Experiments With Different 3d Object Detectorsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Domain Adaptation in LiDAR Semantic Segmentation via Hybrid Learning with Alternating Skip Connections

Corral-Soto

Rochan

et al. 2023

2023 IEEE Intelligent Vehicles Symposium (IV)

Self Cite

View full text Add to dashboard Cite

show abstract

Center transfer for supervised domain adaptation

et al. 2023

View full text Add to dashboard Cite

Domain adaptation (DA) is a popular strategy for pattern recognition and classification tasks. It leverages a large amount of data from the source domain to help train the model applied in the target domain. Supervised domain adaptation (SDA) approaches are desirable when only few labeled samples from the target domain are available. They can be easily adopted in many real-world applications where data collection is expensive. In this study, we propose a new supervision signal, namely center transfer loss (CTL), to efficiently align features under the SDA setting in the deep learning (DL) field. Unlike most previous SDA methods that rely on pairing up training samples, the proposed loss is trainable only using one-stream input based on the mini-batch strategy. The CTL exhibits two main functionalities in training to increase the performance of DL models, i.e., domain alignment and increasing the feature’s discriminative power. The hyper-parameter to balance these two functionalities is waived in CTL, which is the second improvement from the previous approaches. Extensive experiments completed on well-known public datasets show that the proposed method performs better than recent state-of-the-art approaches.

show abstract

Unsupervised Domain Adaptation in LiDAR Semantic Segmentation with Self-Supervision and Gated Adapters

Rochan¹,

Aich²,

Corral-Soto³

et al. 2022

2022 International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

LiDAR few-shot domain adaptation via integrated CycleGAN and 3D object detector with joint learning delay

Cited by 7 publications

References 20 publications

Domain Adaptation in LiDAR Semantic Segmentation via Hybrid Learning with Alternating Skip Connections

Domain Adaptation in LiDAR Semantic Segmentation via Hybrid Learning with Alternating Skip Connections

Center transfer for supervised domain adaptation

Unsupervised Domain Adaptation in LiDAR Semantic Segmentation with Self-Supervision and Gated Adapters

Contact Info

Product

Resources

About