CoChat
Sign Up
Cited by 119Advanced Neural Network ApplicationsAutonomous Vehicle Technology and Safety3D Shape Modeling and Analysis

The DOI matches its Crossref record; 83 of the 91 references it cites were found in a registry or an index.

FocalFormer3D : Focusing on Hard Instance for 3D Object Detection

Open access
Chen, Yilun, Yu, Zhiding, Chen, Yukang, Lan, Shiyi, Anandkumar, Anima, Jia, Jiaya, Alvarez, Jose M.·2023·2023 IEEE/CVF International Conference on Computer Vision (ICCV)
DOI 10.1109/iccv51070.2023.00771
Open this paper in CoChat — free accountOpen PDF

Abstract

False negatives (FN) in 3D object detection, e.g., missing predictions of pedestrians, vehicles, or other obstacles, can lead to potentially dangerous situations in autonomous driving. While being fatal, this issue is understudied in many current 3D detection methods. In this work, we propose Hard Instance Probing (HIP), a general pipeline that identifies FN in a multi-stage manner and guides the models to focus on excavating difficult instances. For 3D object detection, we instantiate this method as FocalFormer3D, a simple yet effective detector that excels at excavating difficult objects and improving prediction recall. FocalFormer3D features a multi-stage query generation to discover hard objects and a box-level transformer decoder to efficiently distinguish objects from massive object candidates. Experimental results on the nuScenes and Waymo datasets validate the superior performance of FocalFormer3D. The advantage leads to strong performance on both detection and tracking, in both LiDAR and multi-modal settings. Notably, FocalFormer3D achieves a 70.5 mAP and 73.9 NDS on nuScenes detection benchmark, while the nuScenes tracking benchmark shows 72.1 AMOTA, both ranking 1st place on the nuScenes LiDAR leaderboard. Our code is available at https://github.com/NVlabs/FocalFormer3D.

Overview

Integrity signals

Registry

Crossref · 10.1109/iccv51070.2023.00771

Cites

83 of 83 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

  • Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

    Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472 · cited 1×

    “This allows for box-level query interaction and iterative box refinements, binding the object queries with sufficient box context through RoIAlign [20, 21] on the bird’s eye view to perform relative bounding box refinements.”
    Method

  • Microsoft COCO: Common Objects in Context

    Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10602-1_48 · Cited by 45,460 · cited 1×

    “Following TransFusion [4], we use a pre-trained ResNet-50 model on COCO [80] and nuIm-age [62] dataset as the image model and freeze its weights during training.”
    Method

  • Focal Loss for Dense Object Detection

    Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Dollar, P. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.324 · Cited by 30,581 · cited 1×

    “Recent research [65–67] has further explored soft-sampling, such as adjusting the loss distribution to mitigate foreground-background imbalance issues.”
    Background

  • Mask R-CNN

    He, K., Gkioxari, G., Dollar, P., & Girshick, R. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.322 · Cited by 29,137

  • MobileNetV2: Inverted Residuals and Linear Bottlenecks

    Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L.-C. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00474 · Cited by 28,755 · cited 1×

    “The generation of multi-stage BEV features is accomplished in a cascade manner using a lightweight inversed residual block [68] between stages.”
    Method

  • Fast R-CNN

    Girshick, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.169 · Cited by 28,105

  • End-to-End Object Detection with Transformers

    Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., & Zagoruyko, S. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58452-8_13 · Cited by 18,285 · cited 2×

    “To enhance the efficiency of object query processing, we employ deformable attention [17] instead of computationally intensive modules such as cross attention [59] or box attention [70].”
    BackgroundMethod

  • Are we ready for autonomous driving? The KITTI vision benchmark suite

    Geiger, A., Lenz, P., & Urtasun, R. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248074 · Cited by 14,897 · cited 1×

    “…detected objects ( Positive predictions) at k -th stage are We are then allowed to classify the ground-truth objects according to their assigned candidates: where an object matching metric σ ( · , · ) (e.g. Intersection over Union [60, 61] and center distance [62]) and a prede-fined threshold η .”
    Background

  • PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric SpacePreprint

    Qi, C. R., Yi, L., Su, H., & Guibas, L. J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1706.02413 · Cited by 10,099 · cited 1×

    “…LiDAR-based [1– 3, 12, 13, 22–29], or Camera-based [30–37], or Multi-Modal [4–8 , 38– 45] 3D object detectors Different from dense detectors, point-based 3D detectors [11, 50–52] directly process point clouds via Point-Net [53, 54] and perform grouping or predictions on the sparse representations.”
    Background

  • PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

    Charles, R., Su, H., Mo, K., & Guibas, L. J. (2017) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2017.16 · Cited by 10,029 · cited 1×

    “…LiDAR-based [1– 3, 12, 13, 22–29], or Camera-based [30–37], or Multi-Modal [4–8 , 38– 45] 3D object detectors Different from dense detectors, point-based 3D detectors [11, 50–52] directly process point clouds via Point-Net [53, 54] and perform grouping or predictions on the sparse representations.”
    Background

Show all 91 references

Research works printed but not linked: 4

  1. Roddick, Orthographic feature transform for monocular 3d object detection
  2. 10.1007/978-3-031-20077-9_1 · DOI · Paper page
  3. 10.1109/tpami.2023.3333838/mm1 · DOI
  4. Zhou, End-to-end multi-view fusion for 3d object detection in lidar point clouds, 923

Cited by

Sorted by citations

See all 119

Related

Metadata

Authors
Chen, Y., Yu, Z., Chen, Y., Lan, S., Anandkumar, A., Jia, J., & Alvarez, J. M.
Affiliations
  • Chinese University of Hong Kong (HK)
  • Nvidia (United Kingdom) (GB)
  • Caltech
  • NVIDIA
ORCID
Venue
2023 IEEE/CVF International Conference on Computer Vision (ICCV)
Publisher
IEEE
Published
2023-10-01
Pages
8360-8371
DOI
10.1109/iccv51070.2023.00771
OpenAlex
W4390872444
Semantic Scholar
ef80a308b548f582ac52419dffaa48f53db8a4f8
Cited by
119
References
91
Citations by year
2026: 35 · 2025: 53 · 2024: 30 · 2023: 1
FWCI
12.73
Type
conference paper
Language
en
Open access
Licence
Locations
  • IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
  • arXiv (Cornell University) (submittedVersion, cc-by-nc-nd)
  • Rare & Special e-Zone (The Hong Kong University of Science and Technology) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
  • Physical Sciences › Computer Science › Computer Graphics and Computer-Aided Design › 3D Shape Modeling and Analysis
Keywords
3D object detection, false negative reduction, LiDAR, autonomous driving, object tracking, multimodal detection, nuScenes, Waymo

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review