Just Add Geometry: Gradient-Free Open-Vocabulary 3D Detection Without Human-in-the-Loop

Atharv Goel; Mehar Khurana

REVIEW 4 major objections 4 minor 26 references

Just Add Geometry: Gradient-Free Open-Vocabulary 3D Detection Without Human-in-the-Loop

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that arbitrary natural-language classes can be localized in 3D with off-the-shelf 2D foundation models and pure geometry, no 3D labels or training needed.

desk verdict A plausible training-free 2D-to-3D lifting recipe with a real benchmark, but the evaluation is thin and the best row quietly uses the dataset-specific priors the paper says it avoids. read the letter →

arxiv 2507.13363 v1 pith:K2PCLZR6 submitted 2025-07-06 cs.CV cs.AI

Atharv Goel , Mehar Khurana This is my paper · ORCID

classification cs.CVcs.AI

keywords open-vocabulary3Dobjectdetectiontraining-free2DfoundationmodelsgeometricboxinflationDBSCANoutlierfilteringrotatingcalipersnuScenesmonocularpseudo-depth

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Open-vocabulary 3D detection usually needs 3D training data or a learned 3D feature extractor. This paper claims to sidestep both: a 2D vision-language detector proposes boxes for arbitrary text prompts, a segmentation model refines them, and camera geometry lifts the resulting masks into 3D point clouds from which an untrained geometric routine recovers 3D bounding boxes. The system is entirely training-free, needs no 3D annotations, and works with either LiDAR or monocular pseudo-depth. On nuScenes mini it reports 29.94% mAP with a medoid-plus-shape-priors variant, close to a prior HD-map-based baseline, and on a new fog-augmented RGB-only benchmark, Pseudo-nuScenes, it reports 12–16% mAP. The paper positions this as the first fully training-free, open-vocabulary 3D detection pipeline built purely from 2D foundation models and geometry.

What carries the argument

The load-bearing mechanism is the 2D-to-3D lifting and inflation loop. GroundingDINO turns a natural-language prompt into 2D boxes; SAM turns each box into an instance mask; known camera intrinsics and extrinsics back-project the pixels covered by the mask onto either LiDAR points or UniDepth pseudo-depth points; DBSCAN keeps the densest cluster to remove projection noise; and a medoid center plus a ground-plane oriented bounding box from Rotating Calipers, with height taken from vertical point extent, yields the final 3D box. The optional shape priors and lane geometry are handcrafted anchors used only in the best-performing variant. What this machinery does is replace learned 3D feature extraction with pure geometry: every stage is deterministic and training-free.

What would settle it

On nuScenes frames where an object is partially occluded or truncated, compare the predicted box against the full annotated box and plot center and size error against the fraction of the object visible in the mask; if errors spike sharply when the visible fraction is small, the coverage assumption fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that an off-the-shelf 2D open-vocabulary detector, combined with a segmentation mask and a classical 3D inflation step, is sufficient to localize arbitrary natural-language classes in 3D without any 3D supervision. The discovery is an application result: mature 2D vision-language models already encode enough object knowledge that their 2D outputs can be turned into usable 3D boxes by projection and geometry alone. The strongest reported evidence is 29.94% mAP on nuScenes mini using the medoid of back-projected LiDAR points as center, with handcrafted lane geometry and shape priors for box size, compared with 34.31% for the rotating-calipers variant of the HD-map-based baseline; replacing LiDAR with UniDepth pseudo-depth gives 12–16% mAP under synthetic fog. The paper does not claim to beat trained 3D detectors; it claims that the ceiling of 2D foundation models transfers to 3D with no additional learning.

Load-bearing premise

The visible 3D points inside a mask, after the densest cluster is kept, cover the object's full extent, so the medoid center and calipers box can match the true annotated box.

Editorial extensions

If this is right

Any class expressible in a 2D open-vocabulary prompt becomes a 3D detection target, so novel categories can be queried at inference time with no 3D class list.
The gap to the HD-map baseline is modest (29.94% vs 34.31% mAP) despite using no maps or dataset priors, so pure geometric inflation is a viable substitute for map-based priors in LiDAR settings.
Pseudo-depth makes the pipeline fully RGB-only, and even under synthetic fog the method retains 12–16% mAP, indicating that monocular depth can stand in for LiDAR in constrained settings.
DBSCAN outlier removal is not a detail: without it the rotating-calipers-only variant drops from 21.94% to 1.30% mAP, so density filtering is what keeps back-projected noise from destroying boxes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A testable extension follows from the pipeline's construction: improving 2D open-vocabulary recall should improve 3D mAP nearly one-for-one, because a 3D proposal can only exist where the 2D detector first found the object.
A control experiment on the new benchmark would isolate error sources: running the same RGB-only pipeline on clear-weather nuScenes images versus Pseudo-nuScenes fog images would separate depth-estimation noise from fog degradation.
Because labels and confidence scores are inherited from 2D detections, the system's semantics are only as good as the 2D detector; a spurious 2D box appears as a spurious 3D box with no learned module to correct it.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Let me save you the read. This paper shows that GroundingDINO + SAM + DBSCAN + rotating calipers can produce 3D boxes for natural-language queries without any training, and it introduces Pseudo-nuScenes, a fog-augmented RGB-only benchmark. The core idea is alive: it works well enough on nuScenes mini to be a useful proof-of-concept, and the code and benchmark could be genuinely useful to people studying 2D-to-3D lifting.

The good parts are real. The DBSCAN ablation (1.3% to 21.9% mAP) is a nice illustration of why noise filtering matters. The pseudo-depth experiments at 12-16% mAP honestly show how much recall you lose when you drop LiDAR. The writing is clear and the pipeline is easy to follow.

The soft spots are in proportion to the claims. The evaluation is 10 validation scenes and 5 classes, with no error bars. The comparison is only with 3D For Free, and their best numbers are 5-14 points lower, so "competitive" is a stretch. There is no comparison with any open-vocabulary 3D baseline, trained or otherwise. More importantly, the paper's Section 2 summary says "no dataset-specific priors," but the best table row uses lane geometry and handcrafted shape priors. That contradiction needs to be resolved. The rotating-calipers-only version drops to 21.94%, so the "no priors" version is significantly weaker. The geometric assumption that the visible points inside a SAM mask cover the object's full extent is unverified; occlusion and truncation will systematically bias centers and sizes. The paper doesn't stratify by visibility or report per-class results, so it can't quantify this. That's a real limitation, but not a fatal one: the method is best described as a visible-surface box estimator, and the claim of general open-vocabulary 3D detection should be softened. References have duplicates and a couple of author lists look wrong; a referee should flag that.

This deserves a serious referee. It's a solid proof-of-concept with a benchmark others might use. I would send it to peer review, but expect heavy revision: fix the priors contradiction, add error bars and per-visibility breakdown, compare to at least one actual open-vocab 3D detector, and clean up the references. The central idea holds; it just needs to be positioned honestly.

Referee Report

4 major / 4 minor

Summary. The paper proposes a training-free, open-vocabulary 3D object detection pipeline that combines off-the-shelf 2D open-vocabulary detectors (GroundingDINO), SAM instance segmentation, back-projection of LiDAR or monocular pseudo-depth into 3D, DBSCAN noise filtering, and classical geometric fitting (medoid centering and rotating calipers) to produce 3D bounding boxes without any 3D-annotated training data. The authors also introduce Pseudo-nuScenes, a fog-augmented RGB-D variant of nuScenes mini, and report 29.94% mAP on nuScenes mini and 12–16% mAP on Pseudo-nuScenes, finding that their best variant is competitive with the older 3D For Free baseline while remaining training-free.

Significance. If validated, the core idea is attractive and practically relevant: it shows that mature 2D vision-language models plus geometric post-processing can produce open-vocabulary 3D boxes without 3D supervision, which is a meaningful step beyond trained 3D open-vocabulary detectors. The pipeline is modular, the components are off-the-shelf, and the authors state they release code. The evaluation is not circular, since the 3D boxes are produced by external 2D models and geometry rather than test-label supervision. However, the current evidence is too thin to support the headline claim of 'competitive' open-vocabulary 3D detection: the evaluation uses only 10 validation scenes and 5 classes, has no error bars or per-class breakdown, does not compare with modern open-vocabulary 3D baselines, and the best-performing variant relies on lane geometry and shape priors that contradict the paper's stated absence of dataset-specific priors. The paper is an interesting position piece, but the central claim needs substantially more support.

major comments (4)

[Section 3.4 and Section 5.2] The geometric inflation step assumes that the back-projected points inside each SAM mask, after DBSCAN keeps only the densest cluster, capture the full geometric extent of the object. The medoid center and the rotating-calipers box are computed from the visible point set, so under occlusion or truncation the box is biased toward the visible side and is systematically smaller than the ground-truth box. Driving scenes frequently contain such cases, and nuScenes provides visibility attributes that would allow a stratified evaluation. The manuscript does not provide any such stratification, so the reported 29.94% mAP cannot be interpreted as evidence that the method handles occluded or truncated objects. Please add an experiment that reports performance separately for visible, occluded, and truncated objects, or clearly state and quantify this limitation.
[Section 5.3, Table 1, Section 6.1] The paper repeatedly claims that the method uses 'no dataset-specific priors' (e.g., Section 2, Section 6.1), but the best-performing row in Table 1 is 'Medoid + Lane geometry + shape priors' (29.94%). Lane geometry and per-class shape priors are dataset-specific priors by construction, and their definitions or values are not provided anywhere in the manuscript. This makes the headline result unreproducible and weakens the 'training-free and prior-free' claim. Please either report the fully prior-free variant (e.g., medoid + rotating calipers without lane geometry and without class-specific shape priors) as the main claim, or explicitly list and justify all priors used.
[Section 6.1] The claim that 'our method featuring the 3D Rotating Calipers strategy yields improved orientation estimates, with a lower mAOE (1.045 vs. 1.144), surpassing the baseline inflation methods [21]' is contradicted by Table 1. The value 1.045 appears in the row 'Medoid + Lane geometry + shape priors', not in a rotating-calipers row, and the 3D For Free baseline with HD maps reports mAOE 0.90, which is lower than 1.045. This comparison should be corrected, and the conclusion about orientation accuracy should be re-evaluated.
[Section 5.1, Tables 1 and 2] The entire evaluation uses only two nuScenes validation sequences (10 scenes total) and 5 classes, with no error bars, no per-class results, and no comparison with the open-vocabulary 3D detection methods discussed in Section 2 (e.g., CLIP-FO3D, Lu et al., FSD). mAP computed on 10 scenes is high-variance, and the absence of any modern open-vocabulary baseline makes the 'competitive localization performance' claim unsupported. Please enlarge the evaluation (at least the full nuScenes validation set), provide error bars or per-sequence results, and compare with at least one recent open-vocabulary 3D detector.

minor comments (4)

[Section 4.1] The fog model is defined with parameters beta and A, but the values of beta are never specified and A is only described as 'set to white'. Without these values, Pseudo-nuScenes is not reproducible.
[Table 1] The row labels are inconsistent and ambiguous: 'Medoid + Lane geometry + shape priors', 'shape priors No', and 'Medoid + Rotating Calipers for orientation, shape' do not clearly indicate which components are active and which are ablated.
[References] References [10] and [11] are duplicate entries for GroundingDINO; the reference list should be de-duplicated and all author names should be checked for typos (e.g., 'V ora' in [1,2]).
[Section 4] The paper states that Pseudo-nuScenes is a 'benchmark', but it is presented as a processed version of the nuScenes mini split with generated pseudo-depth and fog; details about how the dataset will be released (format, license, exact generation script) are missing.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 3D boxes are produced by external 2D detectors, SAM, back-projection, and classical geometry, with no benchmark labels entering the derivation.

full rationale

The derivation chain is not circular. Stage 1 uses an external open-vocabulary 2D detector (Grounding DINO) to produce text-conditioned boxes; Stage 2 uses SAM to segment them; Stage 3 back-projects LiDAR or pseudo-depth points through known camera intrinsics/extrinsics; Stage 4 applies DBSCAN and geometric inflation (medoid centering and rotating calipers) to obtain 3D boxes. None of these stages uses nuScenes 3D ground-truth boxes or labels to compute the output, so the reported mAP values are genuine comparisons against external annotations rather than quantities constructed from the labels. There are no load-bearing self-citations: the references are to external models and methods, and no uniqueness theorem or prior-author result is invoked to force the design. The main concerns in the paper are correctness-related rather than circular: the visible-points-cover-the-object assumption can fail under occlusion and truncation, and the best variant uses handcrafted shape priors and lane geometry that appear benchmark-tuned, with the configuration selected on the same validation set used for reporting. These issues affect generalizability and evaluation validity, but they do not make the prediction equivalent to its inputs by construction. Therefore no circular step is identified and the circularity score is 0.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The pipeline relies on a chain of external models and geometric assumptions; there are no fitted constants in the derivation, but the best reported configuration includes hand-tuned priors, lane geometry and per-class shape priors, whose values are not given. No physical entities are invented; Pseudo-nuScenes is an evaluation artifact, not a new force or particle.

free parameters (5)

GroundingDINO confidence threshold
Controls which 2D proposals become 3D candidates; no value is reported in the paper.
DBSCAN eps and min_samples
Outlier removal relies on these density parameters; the values are not given.
Per-class shape priors = not provided
Handcrafted dimensions used in the best medoid plus shape priors variant; no table of values is supplied.
Lane geometry prior
Map-derived prior used in the best configuration, conflicting with the claim of no dataset-specific priors.
Fog parameters (beta, A) = A set to white, beta unspecified
The fog augmentation in Pseudo-nuScenes requires beta and ambient light values, which are not reported.

assumptions (6)

domain assumption GroundingDINO text-conditioned 2D boxes are reliable proposals for 3D objects.
The whole pipeline inherits 2D detection quality without any verification in 3D, as stated in Stage 1.
domain assumption SAM masks tightly follow object boundaries.
Stage 2 assumes mask pixels correspond to object surfaces; loose masks add outlier points.
standard math Camera intrinsics and extrinsics are known and accurate for back-projection.
Stage 3 projects LiDAR to image; calibration errors would shift masks and 3D points.
domain assumption The densest DBSCAN cluster of masked points contains the target object.
Stage 4 discards all but the densest cluster, assuming background points are not the densest.
domain assumption Rotating calipers on ground-plane points recover true box orientation and size.
Stage 4 fits the minimum-area rectangle to visible 3D points, which fails for partial or occluded objects.
domain assumption UniDepth monocular metric depth is accurate enough for metric 3D reconstruction.
Stage 5 substitutes pseudo-depth for LiDAR; depth errors propagate into box geometry and recall.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Just Add Geometry: Gradient-Free Open-Vocabulary 3D Detection Without Human-in-the-Loop." pith.science (2026). https://pith.science/paper/K2PCLZR6

@misc{pith2026250713363,
  author       = {Pith},
  title        = {Pith review of: Just Add Geometry: Gradient-Free Open-Vocabulary 3D Detection Without Human-in-the-Loop},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K2PCLZR6}},
  note         = {Machine review of arXiv:2507.13363}
}

read the original abstract

Modern 3D object detection datasets are constrained by narrow class taxonomies and costly manual annotations, limiting their ability to scale to open-world settings. In contrast, 2D vision-language models trained on web-scale image-text pairs exhibit rich semantic understanding and support open-vocabulary detection via natural language prompts. In this work, we leverage the maturity and category diversity of 2D foundation models to perform open-vocabulary 3D object detection without any human-annotated 3D labels. Our pipeline uses a 2D vision-language detector to generate text-conditioned proposals, which are segmented with SAM and back-projected into 3D using camera geometry and either LiDAR or monocular pseudo-depth. We introduce a geometric inflation strategy based on DBSCAN clustering and Rotating Calipers to infer 3D bounding boxes without training. To simulate adverse real-world conditions, we construct Pseudo-nuScenes, a fog-augmented, RGB-only variant of the nuScenes dataset. Experiments demonstrate that our method achieves competitive localization performance across multiple settings, including LiDAR-based and purely RGB-D inputs, all while remaining training-free and open-vocabulary. Our results highlight the untapped potential of 2D foundation models for scalable 3D perception. We open-source our code and resources at https://github.com/atharv0goel/open-world-3D-det.

Figures

Figures reproduced from arXiv: 2507.13363 by the authors.

**Figure 1.** Pipeline of our methodology. the 3D domain. This enables open-world 3D object detection in previously unseen environments, offering a scalable path toward 3D perception without the burden of dense 3D annotations. Our key contributions are as follows: • We propose a novel framework for open-vocabulary 3D object detection that requires no manual 3D annotations, leveraging off-the-shelf 2D vision-language models to ge… view at source ↗

**Figure 2.** Effect of DBSCAN-based noise filtering on BEV maps. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Qualitative results on nuScenes (left) and Pseudo [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 21 canonical work pages

[21]

3d for free: Crossmodal transfer learning using hd maps

Benjamin Wilson, Zsolt Kira, and James Hays. 3d for free: Crossmodal transfer learning using hd maps. arXiv preprint arXiv:2008.10592, 2020. 3, 4, 5

arXiv 2008
[1]

Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom

Holger Caesar, Varun Bankiti, Alex H. Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. In CVPR, 2020. 4

work page 2020
[2]

nuscenes: A multimodal dataset for autonomous driv- ing

Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, et al. nuscenes: A multimodal dataset for autonomous driv- ing. In CVPR, 2020. 1, 2, 4

work page 2020
[3]

Glip: Grounded language-image pre-training

Xiao Du, Hong Zhang, Zhen Li, Xiangyang Lin, et al. Glip: Grounded language-image pre-training. In CVPR, 2022. 1, 2

work page 2022
[4]

Are we ready for autonomous driving? the kitti vision benchmark suite

Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In CVPR, 2012. 1, 2

work page 2012
[5]

Lvis: A dataset for large vocabulary instance segmentation

Agrim Gupta, Piotr Dollar, and Ross Girshick. Lvis: A dataset for large vocabulary instance segmentation. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2019. 1

work page 2019
[6]

Berg, Wan-Yen Lo, Piotr Doll ´ar, and Ross Girshick

Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer White- head, Alexander C. Berg, Wan-Yen Lo, Piotr Doll ´ar, and Ross Girshick. Segment anything. arXiv:2304.02643, 2023. 2, 3

arXiv 2023
[7]

Pointpillars: Fast encoders for object detection from point clouds

Alex H Lang, Sourabh V ora, Holger Caesar, Lubing Zhou, Jiong Yang, and Oscar Beijbom. Pointpillars: Fast encoders for object detection from point clouds. In CVPR, 2019. 1, 2

work page 2019

Show all 26 references

[8]

Clip-fo3d: Free open-vocabulary 3d object detection

Zhengxiong Li, Qian Ye, Tianrui Wang, et al. Clip-fo3d: Free open-vocabulary 3d object detection. In ICCV, 2023. 2

2023
[9]

Lawrence Zitnick

Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C. Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision – ECCV 2014 , pages 740–755, Cham,

2014
[11]

Grounding dino: Marrying dino with grounded pre-training for open-set object detection

Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, et al. Grounding dino: Marrying dino with grounded pre-training for open-set object detection. arXiv preprint arXiv:2303.05499, 2023. 1, 2

2023 arXiv
[12]

Clip2scene: Scene-level 3d open-world understanding via vision-language founda- tion models

Yujing Liu, Wenhao Xu, et al. Clip2scene: Scene-level 3d open-world understanding via vision-language founda- tion models. In CVPR, 2023. 2

2023
[13]

Open-vocabulary point-cloud object detection without 3d an- notation

Yuheng Lu, Chenfeng Xu, Xiaobao Wei, Xiaodong Xie, Masayoshi Tomizuka, Kurt Keutzer, and Shanghang Zhang. Open-vocabulary point-cloud object detection without 3d an- notation. Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition, 2023. 1

2023
[14]

Open-vocabulary point-cloud object detection without 3d annotation

Yuheng Lu, Chenfeng Xu, Xiaobao Wei, Xiaodong Xie, Masayoshi Tomizuka, Kurt Keutzer, and Shanghang Zhang. Open-vocabulary point-cloud object detection without 3d annotation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 1190– 1199...

2023
[15]

Unidepth: Universal monocular metric depth estimation

Luigi Piccinelli, Yung-Hsu Yang, Christos Sakaridis, Mattia Segu, Siyuan Li, Luc Van Gool, and Fisher Yu. Unidepth: Universal monocular metric depth estimation. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 2, 3, 4

2024
[16]

Frustum pointnets for 3d object detection from rgb- d data

Charles R Qi, Wei Liu, Chen Wu, Hao Su, and Leonidas J Guibas. Frustum pointnets for 3d object detection from rgb- d data. In CVPR, 2018. 1, 2

2018
[17]

You only look once: Unified, real-time object de- tection

Joseph Redmon, Santosh Divvala, Ross Girshick, and Ali Farhadi. You only look once: Unified, real-time object de- tection. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 779–788, 2016. 1

2016
[18]

Faster r-cnn: Towards real-time object detection with region proposal networks

Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r-cnn: Towards real-time object detection with region proposal networks. IEEE transactions on pattern analysis and machine intelligence, 39(6):1137–1149, 2016. 1

2016
[19]

Scalability in perception for autonomous driving: Waymo open dataset

Pei Sun, Henrik Kretzschmar, et al. Scalability in perception for autonomous driving: Waymo open dataset. In CVPR,
[20]

Fsd: Few-shot object detection in 3d scenes

Yuchen Wang, Qi Zhou, Jianwei Yang, et al. Fsd: Few-shot object detection in 3d scenes. In CVPR, 2023. 2

2023
[22]

Posecnn: A convolutional neural network for 6d object pose estimation in cluttered scenes

Yu Xiang, Tanner Schmidt, Venkatraman Narayanan, and Dieter Fox. Posecnn: A convolutional neural network for 6d object pose estimation in cluttered scenes. In PoseCNN: A Convolutional Neural Network for 6D Object Pose Estima- tion in Cluttered Scenes, 2018. 1

2018
[23]

Ulip: Learning unified representation of language, image and point cloud for 3d understanding

Yujing Xue, Yue Wang, Xingyu Liu, et al. Ulip: Learning unified representation of language, image and point cloud for 3d understanding. In ECCV, 2022. 3

2022
[24]

Open-vocabulary object de- tection using captions

Alireza Zareian, Kevin D Wang, Roozbeh Mottaghi, Ali Farhadi, and Shih-Fu Chang. Open-vocabulary object de- tection using captions. In CVPR, 2021. 1

2021
[25]

Pointclip: Point cloud understanding by clip

Yujing Zhang, Enze Xie, Jiwen Dai, and Zhaoxiang Yu. Pointclip: Point cloud understanding by clip. InCVPR, 2022. 3

2022
[26]

Regionclip: Region- based language-image pretraining

Yiwu Zhong, Jianwei Yang, Pengchuan Zhang, Chun- yuan Li, Noel Codella, Liunian Harold Li, Luowei Zhou, Xiyang Dai, Lu Yuan, Yin Li, et al. Regionclip: Region- based language-image pretraining. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

2022
[2014]

Springer International Publishing. 1

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [21]

3d for free: Crossmodal transfer learning using hd maps

Benjamin Wilson, Zsolt Kira, and James Hays. 3d for free: Crossmodal transfer learning using hd maps. arXiv preprint arXiv:2008.10592, 2020. 3, 4, 5

arXiv 2008

[2] [1]

Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom

Holger Caesar, Varun Bankiti, Alex H. Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. In CVPR, 2020. 4

work page 2020

[3] [2]

nuscenes: A multimodal dataset for autonomous driv- ing

Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, et al. nuscenes: A multimodal dataset for autonomous driv- ing. In CVPR, 2020. 1, 2, 4

work page 2020

[4] [3]

Glip: Grounded language-image pre-training

Xiao Du, Hong Zhang, Zhen Li, Xiangyang Lin, et al. Glip: Grounded language-image pre-training. In CVPR, 2022. 1, 2

work page 2022

[5] [4]

Are we ready for autonomous driving? the kitti vision benchmark suite

Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In CVPR, 2012. 1, 2

work page 2012

[6] [5]

Lvis: A dataset for large vocabulary instance segmentation

Agrim Gupta, Piotr Dollar, and Ross Girshick. Lvis: A dataset for large vocabulary instance segmentation. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2019. 1

work page 2019

[7] [6]

Berg, Wan-Yen Lo, Piotr Doll ´ar, and Ross Girshick

Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer White- head, Alexander C. Berg, Wan-Yen Lo, Piotr Doll ´ar, and Ross Girshick. Segment anything. arXiv:2304.02643, 2023. 2, 3

arXiv 2023

[8] [7]

Pointpillars: Fast encoders for object detection from point clouds

Alex H Lang, Sourabh V ora, Holger Caesar, Lubing Zhou, Jiong Yang, and Oscar Beijbom. Pointpillars: Fast encoders for object detection from point clouds. In CVPR, 2019. 1, 2

work page 2019