Abstract
In this paper, we propose an extended multiple object tracking (MOT) task definition for embodied AI visual exploration research task - multi-class, multi-instance and multi-view object tracking (M3T). The aim of the proposed M3T task is to identify the unique number of objects in the environment, observed on the agent’s way, and visible from far or close view, from different angles or visible only partially. Classic MOT algorithms are not applicable for the M3T task, as they typically target moving single-class multiple object instances in one video and track objects, visible from only one angle or camera viewpoint. Thus, we present the M3T-Round algorithm designed for a simple scenario, where an agent takes 12 image frames, while rotating 360° from the initial position in a scene. We, first, detect each object in all image frames and then track objects (without any training), using cosine similarity metric for association of object tracks. The detector part of our M3T-Round algorithm is compatible with the baseline YOLOv4 algorithm [1] in terms of detection accuracy: a 5.26 point improvement in AP75. The tracker part of our M3T-Round algorithm shows a 4.6 point improvement in HOTA over GMOTv2 algorithm [2], a recent, high-performance tracking method. Moreover, we have collected a new challenging tracking dataset from AI2-Thor [3] simulator for training and evaluation of the proposed M3T-Round algorithm.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bochkovskiy, A., Wang, C.-Y., Liao, H.-Y.M.: YOLOv4: optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934 (2020)
MOT15 results. https://motchallenge.net/results/MOT15/. Accessed 20 Sept 2022
Kolve, E., et al.: AI2-THOR: an interactive 3D environment for visual AI. arXiv preprint arXiv:1712.05474 (2017)
Batra, D., et al.: Rearrangement: a challenge for embodied AI. arXiv preprint arXiv:2011.01975 (2020)
Hall, D., et al.: The robotic vision scene understanding challenge. arXiv preprint arXiv:2009.05246 (2020)
Ciaparrone, G., et al.: Deep learning in video multi-object tracking: a survey. Neurocomputing 381, 61–88 (2020)
Leal-Taixé, L., et al.: MOT challenge 2015: towards a benchmark for multi-target tracking. arXiv preprint arXiv:1504.01942 (2015)
Redmon, J., Ali, F.: YOLOv3: an incremental improvement. arXiv preprint arXiv:1804.02767 (2018)
Girshick, R.: Fast R-CNN. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 1440–1448 (2015)
He, K., et al.: Mask R-CNN. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2961–2969 (2017)
Ren, S., et al.: Faster R-CNN: towards real-time object detection with region proposal networks. In: Advances in Neural Information Processing Systems, vol. 28 (2015)
Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., Zagoruyko, S.: End-to-end object detection with transformers. In: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.-M. (eds.) ECCV 2020. LNCS, vol. 12346, pp. 213–229. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58452-8_13
Zhu, X., et al.: Deformable DETR: deformable transformers for end-to-end object detection. arXiv preprint arXiv:2010.04159 (2020)
Kotar, K., Mottaghi, R.: Interactron: embodied adaptive object detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 14860–14869 (2022)
Yang, J., et al.: Embodied visual recognition. arXiv preprint arXiv:1904.04404 (2019)
Li, A., Liu, L., Wang, K., Liu, S., Yan, S.: Clothing attributes assisted person reidentification. IEEE Trans. Circ. Syst. Video Technol. 25(5), 869–878 (2015)
Farenzena, M., Bazzani, L., Perina, A., Murino, V., Cristani, M.: Person re-identification by symmetry-driven accumulation of local features. In: Proceedings IEEE Conference on Computer Vision and Pattern Recognition, pp. 2360–2367. IEEE (2010)
Zhao, J., et al.: Heterogeneous relational complement for vehicle re-identification. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 205–214 (2021)
Yu, J., et al.: Camera-tracklet-aware contrastive learning for unsupervised vehicle re-identification. In: 2022 International Conference on Robotics and Automation (ICRA), pp. 905–911. IEEE (2022)
Bansal, V., Foresti, G.L., Martinel, N.: Where did i see it? Object instance re-identification with attention. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 298–306 (2021)
Bewley, A., et al.: Simple online and realtime tracking. In: 2016 IEEE International Conference on Image Processing (ICIP), pp. 3464–3468. IEEE (2016)
Wojke, N., Bewley, A., Paulus, D.: Simple online and realtime tracking with a deep association metric. In: 2017 IEEE International Conference on Image Processing (ICIP), pp. 3645–3649. IEEE (2017)
Zhu, X., et al.: ViTT: vision transformer tracker. Sensors 21(16), 5608 (2021)
Sun, P., et al.: TransTrack: multiple object tracking with transformer. arXiv preprint arXiv:2012.15460 (2020)
Meinhardt, T., et al.: TrackFormer: multi-object tracking with transformers. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 8844–8854 (2022)
Zeng, F., et al.: MOTR: end-to-end multiple-object tracking with transformer. arXiv preprint arXiv:2105.03247 (2021)
Savva, M., et al.: Habitat: a platform for embodied AI research. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 9339–9347 (2019)
Luiten, J., et al.: HOTA: A higher order metric for evaluating multi-object tracking. Int. J. Comput. Vis. 129(2), 548–578 (2021). https://doi.org/10.1007/s11263-020-01375-2
Bernardin, K., Stiefelhagen, R.: Evaluating multiple object tracking performance: the clear MOT metrics. EURASIP J. Image Video Process. 1, 1–10 (2008). https://doi.org/10.1155/2008/246309
Ristani, E., Solera, F., Zou, R., Cucchiara, R., Tomasi, C.: Performance measures and a data set for multi-target, multi-camera tracking. In: Hua, G., Jégou, H. (eds.) ECCV 2016. LNCS, vol. 9914, pp. 17–35. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-48881-3_2
Weihs, L., et al.: Visual room rearrangement. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5922–5931 (2021)
Deitke, M., et al.: ProcTHOR: large-scale embodied AI using procedural generation. arXiv preprint arXiv:2206.06994 (2022)
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Piotr Dollár, C., Zitnick, L.: Microsoft COCO: common objects in context. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8693, pp. 740–755. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-10602-1_48
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Khan, M., Abu-Khalaf, J., Suter, D., Rosenhahn, B. (2023). M3T: Multi-class Multi-instance Multi-view Object Tracking for Embodied AI Tasks. In: Yan, W.Q., Nguyen, M., Stommel, M. (eds) Image and Vision Computing. IVCNZ 2022. Lecture Notes in Computer Science, vol 13836. Springer, Cham. https://doi.org/10.1007/978-3-031-25825-1_18
Download citation
DOI: https://doi.org/10.1007/978-3-031-25825-1_18
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-25824-4
Online ISBN: 978-3-031-25825-1
eBook Packages: Computer ScienceComputer Science (R0)