Uncertainty3D: A Lightweight Tri-Dimensional Uncertainty Framework for CNN-Based Active Learning in Object Detection
Source abstract
Abstract
In object detection, annotation cost and computational efficiency are important factors in iterative model improvement under standard benchmark settings. Active learning (AL) addresses this challenge by selecting informative samples for labeling; however, many detection-oriented AL methods incur substantial overhead due to repeated inference (e.g., augmentation-based consistency). This paper introduces Uncertainty3D, a lightweight uncertainty proxy designed for standard CNN-based object detectors. It leverages native pre-NMS predictions to estimate sample informativeness using a single forward pass. We propose a tri-dimensional formulation that captures inconsistencies in position, scale, and category across proposal-consistent predictions. Experiments on PASCAL VOC and MS COCO using representative CNN-based detectors (Faster R-CNN and RetinaNet) show competitive mAP versus representative baselines and about 3–4× faster uncertainty estimation than augmentation-based baselines.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Advanced Neural Network Applications · Adversarial Robustness in Machine Learning · Machine Learning and Algorithms
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Computer Science & Engineering, Beihang University, Beijing 100191, China
China Mobile Research Institute, Beijing 100053, China
China Mobile GBA (Greater Bay Area) Innovation Institute, Guangzhou 510656, China
Funding
National Natural Science Foundation of China
References (39)
Show source-provided references
1. Krizhevsky, A., Sutskever, I., and Hinton, G.E. (2012). ImageNet Classification with Deep Convolutional Neural Networks. Advances in Neural Information Processing Systems, Neural Information Processing Systems Foundation, Inc.
2. Ma, W., Zhang, Z., and Xu, Q. (2020). An Automatic Scheme for Optimizing the Size of Deep Networks. Proceedings of the 3rd International Conference on Signal Processing and Machine Learning, Association for Computing Machinery.10.1145/3432291.3432293 ↗
3. Ma, W., Yuan, X., and Zhu, J. (2023). A Video Image Processing Method for Continuous Object Detection. 8th International Conference on Computer and Communication Systems, IEEE.10.1109/ICCCS57501.2023.10151187 ↗
4. Aggarwal, C.C. (2014). Active Learning: A Survey. Data Classification: Algorithms and Applications, CRC Press.
5. Aghdam, H.H., and Gonzalez-Garcia, A. (2019). Active Learning for Deep Detection Neural Networks. IEEE/CVF International Conference on Computer Vision, IEEE.10.1109/ICCV.2019.00377 ↗
6. Vedaldi, A., Bischof, H., Brox, T., and Frahm, J.M. (2020). Contextual Diversity for Active Learning. Computer Vision—ECCV 2020, Springer International Publishing.10.1007/978-3-030-58610-2 ↗
7. Yu, W., Zhu, S., Yang, T., and Chen, C. (2022). Consistency-Based Active Learning for Object Detection. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), IEEE.10.1109/CVPRW56347.2022.00440 ↗
8. Yoo, D., and Kweon, I.S. (2019). Learning Loss for Active Learning. IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2019.00018 ↗
9. Zhang, Z., Ma, W., Yuan, X., Hao, Y., Guo, M., Tang, H., Zhou, Z., and Yao, Z. (2024). Instance-Aware Uncertainty for Active Learning in Object Detection. 2024 IEEE International Conference on Image Processing (ICIP), IEEE.10.1109/ICIP51287.2024.10647710 ↗
10. Huijser, M., and van Gemert, J.C. (2017). Active Decision Boundary Annotation with Deep Generative Models. IEEE International Conference on Computer Vision, IEEE.10.1109/ICCV.2017.565 ↗
11. Tegen, A., Davidsson, P., and Persson, J.A. (2021). Active Learning and Machine Teaching for Online Learning. 2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA), IEEE.10.1109/ICMLA52953.2021.00197 ↗
12. A Unified Framework for Automatic Distributed Active Learning10.1109/TPAMI.2021.3129793 ↗
13. Online Choice of Active Learning Algorithms
14. Active Learning by Querying Informative and Representative Examples10.1109/TPAMI.2014.2307881 ↗
15. Sener, O., and Savarese, S. (May, January 30). Active Learning for Convolutional Neural Networks: A Core-Set Approach. Proceedings of the International Conference on Learning Representations, Vancouver, BC, Canada.
16. Ash, J.T., Zhang, C., Krishnamurthy, A., Langford, J., and Agarwal, A. (2020, January 26–30). Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds. Proceedings of the International Conference on Learning Representations (ICLR), Addis Ababa, Ethiopia.
17. Beluch, W.H., Genewein, T., Nurnberger, A., and Kohler, J.M. (2018). The Power of Ensembles for Active Learning in Image Classification. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2018.00976 ↗
18. Geifman, Y., Uziel, G., and El-Yaniv, R. (2019, January 6–9). Bias-Reduced Uncertainty Estimation for Deep Neural Classifiers. Proceedings of the International Conference on Learning Representations (ICLR), New Orleans, LA, USA.
19. Wang, K., Yan, X., Zhang, D., Zhang, L., and Lin, L. (2018). Towards Human-Machine Cooperation: Self-supervised Sample Mining for Object Detection. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2018.00173 ↗
20. Jawahar, C., Li, H., Mori, G., and Schindler, K. (2019). Localization-Aware Active Learning for Object Detection. Computer Vision—ACCV 2018, Springer International Publishing.10.1007/978-3-030-20893-6 ↗
21. Dong, J., and Zhang, L. (2023). Multibox Sample Selection for Active Object Detection. 2023 IEEE International Conference on Multimedia and Expo (ICME), IEEE.10.1109/ICME55011.2023.00417 ↗
22. Kendall, A., and Gal, Y. (2017). What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?. Advances in Neural Information Processing Systems (NeurIPS), Neural Information Processing Systems Foundation, Inc.
23. Gal, Y., and Ghahramani, Z. (2016). Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning. 33rd International Conference on Machine Learning (ICML), PMLR.
24. Lakshminarayanan, B., Pritzel, A., and Blundell, C. (2017). Simple and Scalable Predictive Uncertainty Estimation Using Deep Ensembles. Advances in Neural Information Processing Systems (NeurIPS), Neural Information Processing Systems Foundation, Inc.
25. Ovadia, Y., Fertig, E., Ren, J., Nado, Z., Sculley, D., Nowozin, S., Dillon, J.V., Lakshminarayanan, B., and Snoek, J. (2019). Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty under Dataset Shift. Advances in Neural Information Processing Systems (NeurIPS), Neural Information Processing Systems Foundation, Inc.
26. Chen, S., Yang, Y., and Hua, Y. (2023). Semi-Supervised Active Learning for Object Detection. Electronics, 12.10.3390/electronics12020375 ↗
27. Liou, Y.S., Wu, T.H., Yeh, J.F., Chen, W.C., and Hsu, W.H. (2023). MuRAL: Multi-scale Region-Based Active Learning for Object Detection. arXiv.
28. Hekimoglu, A., Brucker, A., Kayali, A.K., Schmidt, M., and Marcos-Ramiro, A. (2023). Active Learning for Object Detection with Non-Redundant Informative Sampling. arXiv.
29. Schubert, M., Riedlinger, T., Kahl, K., and Rottmann, M. (2023). Deep Active Learning with Noisy Oracle in Object Detection. arXiv.10.5220/0012315800003660 ↗
30. Na, J., and De-Silva, V. (2023). Uncertainty Aware Active Learning for Reconfiguration of Pre-Trained Deep Object-Detection Networks for New Target Domains. 2023 IEEE IAS Global Conference on Emerging Technologies (GlobConET), IEEE.10.1109/GlobConET56651.2023.10149954 ↗
31. Yang, C., Huang, L., and Crowley, E.J. (2024). Plug and Play Active Learning for Object Detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE.10.1109/CVPR52733.2024.01684 ↗
32. Yuan, T., Wan, F., Fu, M., Liu, J., and Xu, S. (2021). Multiple Instance Active Learning for Object Detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE.10.1109/CVPR46437.2021.00529 ↗
33. Feng, Z., Zhang, S., Takezoe, R., Hu, W., Chandraker, M., Li, L.J., Narayanan, V.K., and Wang, X. (2022). ALBench: A Framework for Evaluating Active Learning in Object Detection. arXiv.
34. Riedlinger, T., Schubert, M., Kahl, K., Gottschalk, H., and Rottmann, M. (2022). Towards Rapid Prototyping and Comparability in Active Learning for Deep Object Detection. arXiv.
35. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks10.1109/TPAMI.2016.2577031 ↗
36. Focal Loss for Dense Object Detection10.1109/TPAMI.2018.2858826 ↗
37. Houlsby, N., Huszár, F., Ghahramani, Z., and Lengyel, M. (2011). Bayesian Active Learning for Classification and Preference Learning. arXiv.
38. Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., and Zitnick, C.L. (2014). Microsoft COCO: Common Objects in Context. European Conference on Computer Vision, Springer International Publishing.10.1007/978-3-319-10602-1_48 ↗
39. The PASCAL Visual Object Classes (VOC) Challenge10.1007/s11263-009-0275-4 ↗