Pith. sign in

REVIEW 2 major objections 5 minor 52 references

Class-Incremental Motion Forecasting

T0 review · 2 major / 5 minor · reviewed 2026-07-15 · grok-4.5

Pith's one-line read A motion forecaster can learn new road-agent classes over time from cameras without forgetting old ones.

desk verdict Solid first systems paper on class-incremental end-to-end camera forecasting; the task framing and pipeline are real, the PL-fidelity gap is real but not fatal. read the letter →

arxiv 2603.09420 v3 pith:R3PQHQTF submitted 2026-03-10 cs.CV cs.AIcs.RO

classification cs.CVcs.AIcs.RO
keywords class-incrementallearningmotionforecastingend-to-endautonomousdrivingcatastrophicforgettingpseudo-labelingexperiencereplayopen-worldperceptioncamera-basedtrajectoryprediction
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

Autonomous vehicles need to predict where other agents will go, but real roads keep introducing new object types—scooters, new vehicles—while perception is imperfect and full re-annotation of past data is impractical. This paper defines class-incremental (open-world) motion forecasting: the model must predict future trajectories directly from multi-view camera images while new semantic classes arrive in successive steps, with labels only for the current subset. The authors introduce OMEN, which generates motion pseudo-labels for previously learned classes from the old model’s future detections, filters false positives by matching 3D keypoints to open-vocabulary 2D masks, and replays past sequences chosen by high variance in motion-query features. On nuScenes and Argoverse 2 the method retains accuracy on early classes, adapts to new ones, transfers zero-shot to real driving, and extends to class-incremental planning. A sympathetic reader cares because this removes the closed-world barrier that forces full re-training whenever the object taxonomy grows.

What carries the argument

VLM-guided pseudo-label generation plus variance-based sequence replay: the previous model’s detections at future frames become motion pseudo-labels; they are kept only if a majority of projected 3D keypoints fall inside open-vocabulary 2D instance masks of the same class; past sequences with highest sum-of-squared deviations of motion queries from the class mean are stored in a fixed-size replay buffer.

What would settle it

Run the full per-class incremental protocol on nuScenes with the VLM filter disabled or replaced by a weaker detector; if mAP_f and non-linear AP_f on the first classes (car, pedestrian) collapse toward the forgetting baseline while joint-training remains high, the pseudo-label quality claim fails.

Watch

Extended reading notes

Core claim

OMEN is the first end-to-end class-incremental motion forecasting framework: when new classes are introduced sequentially with labels only for the current subset, it preserves forecasting performance on previously learned classes while adapting to novel ones by combining future-detection pseudo-labels, VLM-based 3D-to-2D keypoint filtering, and variance-based sequence replay.

Load-bearing premise

That pseudo-labels built from the previous model’s future detections and filtered by open-vocabulary masks are accurate and calibrated enough to stand in for ground truth on all old classes without ever seeing the original labeled data again.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. The paper formalizes open-world / class-incremental end-to-end motion forecasting: future trajectories are predicted directly from multi-view camera images while new semantic classes are introduced sequentially, with ground-truth labels available only for the current subset Ci and no access to the original full training set. It proposes OMEN, built on SparseDrive, that (i) generates motion pseudo-labels for previously seen classes by composing 3D detections of the prior model Φi−1 at future frames (Eq. 1), (ii) filters false positives / over-confident proposals via 3D-to-2D keypoint majority voting against Grounded-SAM-2 instance masks (Eq. 2), and (iii) maintains a fixed-size sequence replay buffer selected by per-class variance of latent motion queries (Eqs. 4–5). Evaluations on nuScenes (per-class 7-step and group 3-step splits) and Argoverse 2 (20-2 overlapping) show that OMEN retains mAPf / EPA on old classes better than forgetting, plain pseudo-labeling and adapted CL-DETR baselines while approaching a joint-training upper bound; ablations isolate the contributions of future-detection PLs, VLM filtering and variance replay; the framework is extended to open-loop ego planning (Table IV) and demonstrated qualitatively zero-shot on real-vehicle data. Code is promised.

Significance. The work identifies a practically relevant gap—closed-world forecasting models cannot cheaply absorb novel classes (e-scooters, etc.) under storage and re-annotation constraints—and supplies the first end-to-end camera-based solution together with a reproducible experimental protocol (multiple incremental splits, joint upper bound, strong continual-detection baseline). The combination of future-detection pseudo-labels, open-vocabulary VLM filtering and motion-query variance replay is novel for this task; public code and multi-dataset results (including planning transfer) make the contribution usable by the community. If the retention numbers hold under stricter PL-quality scrutiny, the method offers a concrete path toward continual adaptation of the full perception–prediction–planning stack.

major comments (2)
  1. [§III-B, Tables I–III] §III-B (Eqs. 1–2) and the central retention claims of Tables I–III: preservation of old-class forecasting is attributed to VLM-filtered future-detection pseudo-labels that replace ground truth for C0…Ci−1 on Di. The manuscript acknowledges rising confidence / residual false positives and keeps unmatched proposals at zero loss weight, yet reports no precision, recall, calibration or AP of the filtered PLs against the original GT of those classes. Without this quantification it remains unclear how much of the observed gap closure (e.g., car mAPf 29.95 vs joint 32.75) is produced by high-fidelity PLs versus the small variance replay buffer alone; an explicit PL-quality table (or oracle-PL ablation) is needed to make the “mitigates catastrophic forgetting” claim fully load-bearing.
  2. [Abstract, §III-D, Table IV] Abstract and §I claim that the framework “extends naturally to open- and closed-loop end-to-end class-incremental planning on nuScenes and NeuroNCAP.” The body (§III-D, Table IV) only presents open-loop L2 / collision numbers on the nuScenes per-class split; no closed-loop results and no NeuroNCAP experiments appear. Either the missing experiments must be added or the abstract / contribution list must be revised to match the delivered content.
minor comments (5)
  1. [Title / Abstract] Title uses “Class-Incremental Motion Forecasting” while the body consistently speaks of “open-world motion forecasting”; align terminology for clarity.
  2. [§H, Abstract] §H Limitations correctly notes that old classes need not be fully absent and that GT is still required for novel classes; these caveats should also be reflected briefly in the abstract so readers do not over-interpret the “no original training set” claim.
  3. [Table V] Table V ablation rows are dense; adding a short textual summary of the incremental gains (future-det PL o +VLM o +var buffer) would improve readability.
  4. [§G] Real-world zero-shot results (Fig. 4, §G) are purely qualitative; a short note on domain-gap statistics (camera FOV, mounting height, country) would help readers gauge transfer strength.
  5. [Eq. (1)] Minor notation: the homogeneous transform in Eq. (1) is written T au o t but later text uses T au o t inconsistently; fix subscript style.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; empirical incremental-learning method whose reported metrics are independent of the training pseudo-labels and replay scores.

full rationale

The paper introduces a class-incremental end-to-end motion-forecasting setting and an engineering pipeline (future-detection pseudo-labels composed via Eq. 1, VLM keypoint-mask majority filter via Eq. 2, variance-based sequence replay via Eqs. 4-5). All of these are training-time mechanisms. The central claims (retention of old-class mAPf / EPA while adapting to new classes; Tables I-III) are measured on held-out validation sets against an independent joint-training upper bound, a pure-forgetting baseline, and adapted CL-DETR, using standard external metrics that do not algebraically reduce to the pseudo-label or replay constructions. No uniqueness theorem, fitted identity, or self-citation is load-bearing for the reported numbers; prior self-citations (SparseDrive backbone, etc.) supply architecture, not the incremental result. The acknowledged limitations on pseudo-label fidelity are correctness risks, not circular reductions. The derivation chain is therefore self-contained against external benchmarks.

Assumptions & free parameters 5 free parameters · 5 assumptions · 3 invented entities

The central empirical claim rests on standard DETR-style end-to-end forecasting assumptions, a handful of hand-chosen training/replay thresholds, and the reliability of an external open-vocabulary segmenter—not on new physical entities. Free parameters control pseudo-label acceptance and memory; domain axioms include query-identity tracking and imperfect perception; invented constructs are the OMEN pipeline pieces (variance replay score and 3D-to-2D voting rule).

free parameters (5)
  • pseudo-label confidence threshold θ = 0.3
    Fixed at 0.3 for retaining detection proposals before VLM matching; directly controls FP/FN trade-off in old-class labels.
  • replay buffer size N_replay = 30 sequences
    Fixed at 30 sequences (ablated at 6/12/30); divides evenly across past classes and is a hand-chosen memory budget the forgetting claim depends on.
  • keypoint majority threshold = > 1/2
    Keep proposal if >1/2 of projected keypoints fall in a same-class VLM mask (Eq. 2); hand-set voting rule for filtering.
  • stage-wise learning rates and epoch counts = as in §IV-C
    Base vs incremental LRs (2e-4/1e-4 halved later), 50+10 epochs (nuScenes) / 10+2 (AV2), ~25% replay in stage 2; chosen training schedule affecting plasticity/forgetting.
  • per-class sequence subsample size = 300 sequences/class
    300 sequences per class on nuScenes per-class split (or all if fewer); defines the incremental data regime the results are measured under.
assumptions (5)
  • domain assumption End-to-end DETR-like models with query propagation provide consistent object identities across time so future detections can be chained into motion pseudo-labels without external association.
    Stated in §III-B.1 as the basis for composing trajectories from future detection centers via T_τ→t.
  • domain assumption Grounded SAM 2 / Grounding DINO open-vocabulary masks are sufficiently accurate on driving images to filter false-positive 3D proposals for previously learned classes.
    Core of §III-B.2; without this, pseudo-label quality and confidence calibration claims fail.
  • domain assumption A small fixed exemplar buffer of past sequences plus pseudo-labels on new data is an acceptable substitute for retaining the full historical labeled dataset under storage constraints.
    Problem formulation §III-A and experience replay §III-C; standard continual-learning premise made load-bearing here.
  • domain assumption Hungarian matching between predictions and ground truth (or pseudo-labels) correctly associates motion queries for variance scoring and training.
    Used in §III-C to obtain q_j,motion per instance for the replay score s_k,c.
  • domain assumption Standard supervised detection and motion losses from SparseDrive remain valid when some labels are pseudo-labels with zero weight for unmatched proposals.
    Training protocol §IV-C and matching description in §III-B.2.
invented entities (3)
  • OMEN class-incremental forecasting pipeline
    purpose: Unify future-detection pseudo-labeling, VLM filtering, and sequence replay for open-world camera-based motion forecasting.
    Named system introduced as the first method for this setting; evaluated as a whole against baselines.
  • Motion-query variance sequence score s_k,c
    purpose: Select replay sequences with informative motion patterns via sum of squared deviations of motion queries from the class mean query.
    Defined in Eqs. 4–5 §III-C; not a physical entity but a new selection statistic specific to this paper.
  • 3D-to-2D keypoint voting match to VLM masks
    purpose: Accept or reject 3D pseudo-label proposals by majority projection into open-vocabulary instance masks.
    Eq. 2 §III-B.2; mechanism invented for this pipeline though built from existing keypoint sampling and Grounded SAM 2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Class-Incremental Motion Forecasting." pith.science (2026). https://pith.science/paper/R3PQHQTF

@misc{pith2026260309420,
  author       = {Pith},
  title        = {Pith review of: Class-Incremental Motion Forecasting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R3PQHQTF}},
  note         = {Machine review of arXiv:2603.09420}
}
read the original abstract

Motion forecasting enables autonomous vehicles to anticipate scene evolution by predicting the future trajectories of dynamic agents. However, existing approaches typically assume a closed-world setting with a fixed object taxonomy and access to high-quality perception, limiting their applicability in the real world where perception is imperfect, and new object classes may emerge over time. In this work, we introduce class-incremental motion forecasting, a novel setting in which new object classes are sequentially introduced over time and future object trajectories are predicted directly from camera images. We propose the first end-to-end framework for this setting, which adapts to newly introduced classes while mitigating catastrophic forgetting of previously learned ones. Our method generates motion forecasting pseudo-labels for known classes and matches them with 2D instance masks from an open-vocabulary segmentation model. This 3D-to-2D keypoint voting mechanism filters inconsistent and overconfident predictions, while a query feature variance-based replay strategy samples informative past sequences to preserve prior knowledge. Extensive evaluations on nuScenes and Argoverse 2 show that our approach successfully preserves performance on known classes while effectively adapting to novel ones. We further demonstrate zero-shot transfer to real-world driving and show that the framework extends naturally to open- and closed-loop end-to-end class-incremental planning on nuScenes and NeuroNCAP. Code and models will be made publicly available at https://omen.cs.uni-freiburg.de.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

52 extracted references · 8 linked inside Pith

  1. [1]

    Lifelong vehicle trajectory prediction frame- work based on generative replay.IEEE Transactions on Intelligent Transportation Systems, 24(12):13729–13741, 2023

    Peng Bao, Zonghai Chen, Jikai Wang, Deyun Dai, and Hao Zhao. Lifelong vehicle trajectory prediction frame- work based on generative replay.IEEE Transactions on Intelligent Transportation Systems, 24(12):13729–13741, 2023

  2. [2]

    nuscenes: A multimodal dataset for autonomous driving

    Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Giancarlo Baldan, and Oscar Beijbom. nuscenes: A multimodal dataset for autonomous driving. InIEEE Conf. on Computer Vision and Pattern Recognition, pages 11621–11631, 2020

  3. [3]

    The importance of prior knowledge in precise multimodal prediction

    Sergio Casas, Cole Gulino, Simon Suo, and Raquel Urtasun. The importance of prior knowledge in precise multimodal prediction. In2020 IEEE/RSJ international conference on intelligent robots and systems (IROS), pages 2295–2302. IEEE, 2020

  4. [4]

    Modeling the background for incremental learning in semantic segmentation

    Fabio Cermelli, Massimiliano Mancini, Samuel Rota Bulo, Elisa Ricci, and Barbara Caputo. Modeling the background for incremental learning in semantic segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 9233– 9242, 2020

  5. [5]

    Traphic: Trajectory prediction in dense and heterogeneous traffic using weighted interactions

    Rohan Chandra, Uttaran Bhattacharya, Aniket Bera, and Dinesh Manocha. Traphic: Trajectory prediction in dense and heterogeneous traffic using weighted interactions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8483–8492, 2019

  6. [6]

    A new knowledge distillation for incremental object detection

    Li Chen, Chunyan Yu, and Lvcai Chen. A new knowledge distillation for incremental object detection. In2019 in- ternational joint conference on neural networks (IJCNN), pages 1–7. IEEE, 2019

  7. [7]

    Modeling vehicle interactions via modified lstm models for trajectory prediction.Ieee Access, 7:38287–38296, 2019

    Shengzhe Dai, Li Li, and Zhiheng Li. Modeling vehicle interactions via modified lstm models for trajectory prediction.Ieee Access, 7:38287–38296, 2019

  8. [8]

    Convolutional social pooling for vehicle trajectory prediction

    Nachiket Deo and Mohan M Trivedi. Convolutional social pooling for vehicle trajectory prediction. InProceedings of the IEEE conference on computer vision and pattern recognition workshops, pages 1468–1476, 2018

Show all 52 references
  1. [9]

    Multi-modal tra- jectory prediction of surrounding vehicles with maneuver based lstms

    Nachiket Deo and Mohan M Trivedi. Multi-modal tra- jectory prediction of surrounding vehicles with maneuver based lstms. In2018 IEEE intelligent vehicles symposium (IV), pages 1179–1184. IEEE, 2018

  2. [10]

    Motion forecasting via model-based risk minimization

    Aron Distelzweig, Eitan Kosman, Andreas Look, Faris Janjoˇs, Denesh K Manivannan, and Abhinav Valada. Motion forecasting via model-based risk minimization. InIEEE International Conference on Robotics and Automation (ICRA), pages 7011–7018, 2025

  3. [11]

    Stochasticity in motion: An information-theoretic approach to trajectory prediction

    Aron Distelzweig, Andreas Look, Eitan Kosman, Faris Janjoˇs, J¨org Wagner, and Abhinav Valada. Stochasticity in motion: An information-theoretic approach to trajectory prediction. InIEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 6211–6218, 2025

  4. [12]

    Vectornet: Encoding hd maps and agent dynamics from vectorized representation

    Jiyang Gao, Chen Sun, Hang Zhao, Yi Shen, Dragomir Anguelov, Congcong Li, and Cordelia Schmid. Vectornet: Encoding hd maps and agent dynamics from vectorized representation. InProceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition, pages 11525–11533, 2020

  5. [13]

    Vip3d: End-to- end visual trajectory prediction via 3d agent queries

    Junru Gu, Chenxu Hu, Tianyuan Zhang, Xuanyao Chen, Yilun Wang, Yue Wang, and Hang Zhao. Vip3d: End-to- end visual trajectory prediction via 3d agent queries. In IEEE Conf. on Computer Vision and Pattern Recognition, pages 5496–5506, June 2023

  6. [14]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

  7. [15]

    Taxonomy-aware continual semantic segmentation in hyperbolic spaces for open-world perception.IEEE Robotics and Automation Letters, 10(2):1904–1911, 2024

    Julia Hindel, Daniele Cattaneo, and Abhinav Valada. Taxonomy-aware continual semantic segmentation in hyperbolic spaces for open-world perception.IEEE Robotics and Automation Letters, 10(2):1904–1911, 2024

  8. [16]

    Fiery: Future instance prediction in bird’s-eye view from surround monocular cameras

    Anthony Hu, Zak Murez, Nikhil Mohan, Sof ´ıa Dudas, Jeffrey Hawke, Vijay Badrinarayanan, Roberto Cipolla, and Alex Kendall. Fiery: Future instance prediction in bird’s-eye view from surround monocular cameras. In Proceedings of the IEEE/CVF International Conference on Computer...

  9. [17]

    Planning-oriented autonomous driving

    Yihan Hu, Jiazhi Yang, Li Chen, Keyu Li, Chonghao Sima, Xizhou Zhu, Siqi Chai, Senyao Du, Tianwei Lin, Wenhai Wang, et al. Planning-oriented autonomous driving. In IEEE Conf. on Computer Vision and Pattern Recognition, pages 17853–17862, 2023

  10. [18]

    Multi-modal motion prediction with transformer-based neural network for autonomous driving

    Zhiyu Huang, Xiaoyu Mo, and Chen Lv. Multi-modal motion prediction with transformer-based neural network for autonomous driving. In2022 International Conference on Robotics and Automation (ICRA), pages 2605–2611. IEEE, 2022

  11. [19]

    Perceive, interact, predict: Learning dynamic and static clues for end-to-end motion prediction.arXiv preprint arXiv:2212.02181, 2022

    Bo Jiang, Shaoyu Chen, Xinggang Wang, Bencheng Liao, Tianheng Cheng, Jiajie Chen, Helong Zhou, Qian Zhang, Wenyu Liu, and Chang Huang. Perceive, interact, predict: Learning dynamic and static clues for end-to-end motion prediction.arXiv preprint arXiv:2212.02181, 2022

  12. [20]

    Far3d: Expanding the horizon for surround-view 3d object detection

    Xiaohui Jiang, Shuailin Li, Yingfei Liu, Shihao Wang, Fan Jia, Tiancai Wang, Lijin Han, and Xiangyu Zhang. Far3d: Expanding the horizon for surround-view 3d object detection. InProceedings of the AAAI conference on artificial intelligence, pages 2561–2569, 2024

  13. [21]

    Towards open world object detection

    KJ Joseph, Salman Khan, Fahad Shahbaz Khan, and Vineeth N Balasubramanian. Towards open world object detection. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5830– 5840, 2021

  14. [22]

    Con- tinual learning for motion prediction model via meta- representation learning and optimal memory buffer reten- tion strategy

    DaeJun Kang, Dongsuk Kum, and Sanmin Kim. Con- tinual learning for motion prediction model via meta- representation learning and optimal memory buffer reten- tion strategy. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15438–15448, 2024

  15. [23]

    An energy and gpu-computation efficient backbone network for real-time object detection

    Youngwan Lee, Joong-won Hwang, Sangrok Lee, Yuseok Bae, and Jongyoul Park. An energy and gpu-computation efficient backbone network for real-time object detection. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, pages 0–0, 2019

  16. [24]

    Decode: Domain- aware continual domain expansion for motion prediction

    Boqi Li, Haojie Zhu, and Henry X Liu. Decode: Domain- aware continual domain expansion for motion prediction. arXiv preprint arXiv:2411.17917, 2024

  17. [25]

    End-to- end contextual perception and prediction with interaction transformer

    Lingyun Luke Li, Bin Yang, Ming Liang, Wenyuan Zeng, Mengye Ren, Sean Segal, and Raquel Urtasun. End-to- end contextual perception and prediction with interaction transformer. In2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5784–

  18. [26]

    Learning lane graph representations for motion forecasting

    Ming Liang, Bin Yang, Rui Hu, Yun Chen, Renjie Liao, Song Feng, and Raquel Urtasun. Learning lane graph representations for motion forecasting. InEuropean Conference on Computer Vision, pages 541–556. Springer, 2020

  19. [27]

    Sparse4d v3: Advancing end-to-end 3d detection and tracking.arXiv preprint arXiv:2311.11722, 2023

    Xuewu Lin, Zixiang Pei, Tianwei Lin, Lichao Huang, and Zhizhong Su. Sparse4d v3: Advancing end-to-end 3d detection and tracking.arXiv preprint arXiv:2311.11722, 2023

  20. [28]

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

    Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Qing Jiang, Chunyuan Li, Jianwei Yang, Hang Su, et al. Grounding dino: Marrying dino with grounded pre-training for open-set object detection. In ECCV, pages 38–55, 2024

  21. [29]

    Multi-task incremental learn- ing for object detection.arXiv preprint arXiv:2002.05347, 2020

    Xialei Liu, Hao Yang, Avinash Ravichandran, Rahul Bhotika, and Stefano Soatto. Multi-task incremental learn- ing for object detection.arXiv preprint arXiv:2002.05347, 2020

  22. [30]

    Continual detection transformer for incremen- tal object detection

    Yaoyao Liu, Bernt Schiele, Andrea Vedaldi, and Christian Rupprecht. Continual detection transformer for incremen- tal object detection. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 23799–23808, 2023

  23. [31]

    Multimodal motion prediction with stacked transformers

    Yicheng Liu, Jinghuai Zhang, Liangji Fang, Qinhong Jiang, and Bolei Zhou. Multimodal motion prediction with stacked transformers. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 7577–7586, 2021

  24. [32]

    Cooler: class-incremental learning for appearance-based multiple object tracking

    Zhizheng Liu, Mattia Segu, and Fisher Yu. Cooler: class-incremental learning for appearance-based multiple object tracking. InDAGM German Conference on Pattern Recognition, pages 443–458. Springer, 2023

  25. [33]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. InInternational Conference on Learning Representations, 2019

  26. [34]

    Evidential uncertainty estimation for multi-modal trajectory prediction

    Sajad Marvi, Christoph Rist, Julian Schmidt, Julian Jordan, and Abhinav Valada. Evidential uncertainty estimation for multi-modal trajectory prediction. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 785–792, 2025

  27. [35]

    Wayformer: Motion forecasting via simple & efficient attention networks.arXiv preprint arXiv:2207.05844, 2022

    Nigamaa Nayakanti, Rami Al-Rfou, Aurick Zhou, Kratarth Goel, Khaled S Refaat, and Benjamin Sapp. Wayformer: Motion forecasting via simple & efficient attention networks.arXiv preprint arXiv:2207.05844, 2022

  28. [36]

    Forecasting from lidar via future object detection

    Neehar Peri, Jonathon Luiten, Mengtian Li, Aljo ˇsa O ˇsep, Laura Leal-Taix´e, and Deva Ramanan. Forecasting from lidar via future object detection. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17202–17211, June 2022

  29. [37]

    Grounded sam: Assembling open- world models for diverse visual tasks.arXiv preprint arXiv:2401.14159, 2024

    Tianhe Ren, Shilong Liu, Ailing Zeng, Jing Lin, Kunchang Li, He Cao, Jiayu Chen, Xinyu Huang, Yukang Chen, Feng Yan, et al. Grounded sam: Assembling open- world models for diverse visual tasks.arXiv preprint arXiv:2401.14159, 2024

  30. [38]

    Incremental learning of object detectors without catastrophic forgetting

    Konstantin Shmelkov, Cordelia Schmid, and Karteek Alahari. Incremental learning of object detectors without catastrophic forgetting. InProceedings of the IEEE international conference on computer vision, pages 3400– 3409, 2017

  31. [39]

    Dinov3.arXiv preprint arXiv:2508.10104, 2025

    Oriane Sim ´eoni, Huy V V o, Maximilian Seitzer, Federico Baldassarre, Maxime Oquab, Cijo Jose, Vasil Khalidov, Marc Szafraniec, Seungeun Yi, Micha ¨el Ramamonjisoa, et al. Dinov3.arXiv preprint arXiv:2508.10104, 2025

  32. [40]

    Sparsedrive: End-to-end autonomous driving via sparse scene representation

    Wenchao Sun, Xuewu Lin, Yining Shi, Chuang Zhang, Haoran Wu, and Sifa Zheng. Sparsedrive: End-to-end autonomous driving via sparse scene representation. In Int. Conf. on Robotics & Automation, pages 8795–8801. IEEE, 2025

  33. [41]

    Covio: Online continual learning for visual-inertial odometry

    Niclas V ¨odisch, Daniele Cattaneo, Wolfram Burgard, and Abhinav Valada. Covio: Online continual learning for visual-inertial odometry. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2464–2473, 2023

  34. [42]

    Codeps: Online continual learning for depth estimation and panoptic segmentation

    Niclas V ¨odisch, K ¨ursat Petek, Wolfram Burgard, and Abhinav Valada. Codeps: Online continual learning for depth estimation and panoptic segmentation. InRobotics: Science and Systems (RSS), 2023

  35. [43]

    Parkdiffusion: Heterogeneous multi- agent multi-modal trajectory prediction for automated parking using diffusion models

    Jiarong Wei, Niclas V ¨odisch, Anna Rehr, Christian Feist, and Abhinav Valada. Parkdiffusion: Heterogeneous multi- agent multi-modal trajectory prediction for automated parking using diffusion models. InIEEE/RSJ Interna- tional Conference on Intelligent Robots and Systems (IRO...

  36. [44]

    Parkdiffusion++: Ego intention conditioned joint multi-agent trajectory prediction for automated parking using diffusion models.arXiv preprint arXiv:2602.20923, 2026

    Jiarong Wei, Anna Rehr, Christian Feist, and Abhinav Valada. Parkdiffusion++: Ego intention conditioned joint multi-agent trajectory prediction for automated parking using diffusion models.arXiv preprint arXiv:2602.20923, 2026

  37. [45]

    Argoverse 2: Next generation datasets for self-driving perception and forecasting

    Benjamin Wilson, William Qi, Tanmay Agarwal, John Lambert, Jagjeet Singh, Siddhesh Khandelwal, Bowen Pan, Ratnesh Kumar, Andrew Hartnett, Jhony Kaesemodel Pontes, Deva Ramanan, Peter Carr, and James Hays. Argoverse 2: Next generation datasets for self-driving perception and fo...

  38. [46]

    Motion trajectory prediction based on a cnn-lstm sequential model.Science China Information Sciences, 63(11):212207, 2020

    Guo Xie, Anqi Shangguan, Rong Fei, Wenjiang Ji, Weigang Ma, and Xinhong Hei. Motion trajectory prediction based on a cnn-lstm sequential model.Science China Information Sciences, 63(11):212207, 2020

  39. [47]

    Towards motion forecasting with real-world perception inputs: Are end-to-end approaches competitive? In Int

    Yihong Xu, Lo ¨ıck Chambon, ´Eloi Zablocki, Micka ¨el Chen, Alexandre Alahi, Matthieu Cord, and Patrick P ´erez. Towards motion forecasting with real-world perception inputs: Are end-to-end approaches competitive? In Int. Conf. on Robotics & Automation, pages 18428–18435. IEEE, 2024

  40. [48]

    Multi-view correlation distillation for incremental object detection

    Dongbao Yang, Yu Zhou, Aoting Zhang, Xurui Sun, Dayan Wu, Weiping Wang, and Qixiang Ye. Multi-view correlation distillation for incremental object detection. Pattern Recognition, 131:108863, 2022

  41. [49]

    Vehicle motion prediction at intersections based on the turning intention and prior trajectories model

    Ting Zhang, Wenjie Song, Mengyin Fu, Yi Yang, and Meiling Wang. Vehicle motion prediction at intersections based on the turning intention and prior trajectories model. IEEE/CAA Journal of Automatica Sinica, 8(10):1657– 1666, 2021

  42. [50]

    Bev- erse: Unified perception and prediction in birds-eye-view for vision-centric autonomous driving.arXiv preprint arXiv:2205.09743, 2022

    Yunpeng Zhang, Zheng Zhu, Wenzhao Zheng, Junjie Huang, Guan Huang, Jie Zhou, and Jiwen Lu. Bev- erse: Unified perception and prediction in birds-eye-view for vision-centric autonomous driving.arXiv preprint arXiv:2205.09743, 2022

  43. [51]

    Hivt: Hierarchical vector transformer for multi- agent motion prediction

    Zikang Zhou, Luyao Ye, Jianping Wang, Kui Wu, and Kejie Lu. Hivt: Hierarchical vector transformer for multi- agent motion prediction. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8823–8833, 2022

  44. [52]

    A recurrent neural network solution for predicting driver intention at unsignalized intersections.IEEE Robotics and Automation Letters, 3(3):1759–1764, 2018

    Alex Zyner, Stewart Worrall, and Eduardo Nebot. A recurrent neural network solution for predicting driver intention at unsignalized intersections.IEEE Robotics and Automation Letters, 3(3):1759–1764, 2018. Open-World Motion Forecasting - Supplementary Material - S.1. 3D OBJECT...

Pith tools

Reviewed July 15, 2026 · model on record in the stance chip above.