Scalability in Perception for Autonomous Driving: Waymo Open Dataset

Sun, Pei; Kretzschmar, Henrik; Dotiwalla, Xerxes; Chouard, Aurelien; Patnaik, Vijaysai; Tsui, Paul; Guo, James C. Y.; Zhou, Yin; Chai, Yuning; Caine, Benjamin; Vasudevan, Vijay; Han, Wei; Ngiam, Jiquan; Zhao, Hang; Timofeev, Aleksei; Ettinger, Scott; Krivokon, Maxim; Gao, Amy; Joshi, Aditya; Zhao, Sheng; Cheng, Shuyang; Shlens, Jonathon; Chen, Zhifeng; Anguelov, Dragomir

doi:10.48550/arxiv.1912.04838

Cited by 68 publications

(75 citation statements)

References 24 publications

(24 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result of these contributions, our method greatly outperforms the SoTA on several benchmarks: FT3D [44], stereoKITTI [45], lidarKITTI [22], and semanticKITTI [5], while generalizing to the waymo-open dataset [69] without additional fine-tuning.…”

Section: Introductionmentioning

confidence: 89%

Weakly Supervised Learning of Rigid 3D Scene Flow

Gojčič¹,

Litany²,

Guibas³

et al. 2021

Preprint

View full text Add to dashboard Cite

We propose a data-driven scene flow estimation algorithm exploiting the observation that many 3D scenes can be explained by a collection of agents moving as rigid bodies. At the core of our method lies a deep architecture able to reason at the object-level by considering 3D scene flow in conjunction with other 3D tasks. This object level abstraction, enables us to relax the requirement for dense scene flow supervision with simpler binary background segmentation mask and ego-motion annotations. Our mild supervision requirements make our method well suited for recently released massive data collections for autonomous driving, which do not contain dense scene flow annotations. As output, our model provides low-level cues like pointwise flow and higher-level cues such as holistic scene understanding at the level of rigid objects. We further propose a test-time optimization refining the predicted rigid scene flow. We showcase the effectiveness and generalization capacity of our method on four different autonomous driving datasets. We release our source code and pre-trained models under github.com/zgojcic/Rigid3DSceneFlow.

show abstract

Section: Introductionmentioning

confidence: 89%

Weakly Supervised Learning of Rigid 3D Scene Flow

Gojčič¹,

Litany²,

Guibas³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The closest work to ours is Sun et. al [7], who used confidence-based tracking and employed addition as their update function. However, adding the scores makes the system extra-confident and results in high numbers of false positives.…”

Section: B 3d Multi-object Trackingmentioning

confidence: 99%

“…One function that satisfies this criterion is adding the tracklet and detection scores (equation 4), as was used by [7].…”

Section: Score-updatementioning

confidence: 99%

“…In this case, tracklets consistently matched over time will have high scores, and unmatched tracklets' scores will decline. To the best of our knowledge, [7] was the only work that used confidence-based tracking. They added the detection's and tracklet's scores to update the tracklet's score.…”

Section: Introductionmentioning

confidence: 99%

“…Our functions lead to more stable tracking scores, which eventually lead to better performance. We tested our method on both nuScenes [8] and Waymo [7] datasets with multiple detectors and filtering algorithms. Results showed that the confidence-based method with score refinement consistently provides higher AMOTA and MOTA results than count-based methods with our proposed functions.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Score refinement for confidence-based 3D multi-object tracking

Benbarka¹,

Schröder²,

Zell³

2021

Preprint

View full text Add to dashboard Cite

Multi-object tracking is a critical component in autonomous navigation, as it provides valuable information for decision-making. Many researchers tackled the 3D multiobject tracking task by filtering out the frame-by-frame 3D detections; however, their focus was mainly on finding useful features or proper matching metrics. Our work focuses on a neglected part of the tracking system: score refinement and tracklet termination. We show that manipulating the scores depending on time consistency while terminating the tracklets depending on the tracklet score improves tracking results. We do this by increasing the matched tracklets' score with score update functions and decreasing the unmatched tracklets' score. Compared to count-based methods, our method consistently produces better AMOTA and MOTA scores when utilizing various detectors and filtering algorithms on different datasets. The improvements in AMOTA score went up to 1.83 and 2.96 in MOTA. We also used our method as a late-fusion ensembling method, and it performed better than voting-based ensemble methods by a solid margin. It achieved an AMOTA score of 67.6 on nuScenes test evaluation, which is comparable to other state-of-the-art trackers. Code is publicly available at: https://github.com/cogsys-tuebingen/CBMOT.

show abstract

Improving 3D Object Detection Through Progressive Population Based Augmentation

Cheng¹,

Leng²,

Cubuk

et al. 2020

Computer Vision – ECCV 2020

Self Cite

View full text Add to dashboard Cite

Data augmentation has been widely adopted for object detection in 3D point clouds. All previous efforts have focused on manually designing specific data augmentation methods for individual architectures, however no work has attempted to automate the design of data augmentation in 3D detection problems -as is common in 2D imagebased computer vision. In this work, we present the first attempt to automate the design of data augmentation policies for 3D object detection. We present an algorithm, termed Progressive Population Based Augmentation (PPBA). PPBA learns to optimize augmentation strategies by narrowing down the search space and adopting the best parameters discovered in previous iterations. On the KITTI test set, PPBA improves the StarNet detector by substantial margins on the moderate difficulty category of cars, pedestrians, and cyclists, outperforming all current state-of-the-art single-stage detection models. Additional experiments on the Waymo Open Dataset indicate that PPBA continues to effectively improve 3D object detection on a 20x larger dataset compared to KITTI. The magnitude of the improvements may be comparable to advances in 3D perception architectures and the gains come without an incurred cost at inference time. In subsequent experiments, we find that PPBA may be up to 10x more data efficient than baseline 3D detection models without augmentation, highlighting that 3D detection models may achieve competitive accuracy with far fewer labeled examples.

show abstract

Scalability in Perception for Autonomous Driving: Waymo Open Dataset

Cited by 68 publications

References 24 publications

Weakly Supervised Learning of Rigid 3D Scene Flow

Weakly Supervised Learning of Rigid 3D Scene Flow

Score refinement for confidence-based 3D multi-object tracking

Improving 3D Object Detection Through Progressive Population Based Augmentation

Contact Info

Product

Resources

About