Abstract
This paper presents an unsupervised object detection system which can offline-learn generic visual features via Siamese neural network, yet is able to learn new object classes at run-time with a prototype learning approach applied on the latent representations. The operating requirements of this system feature bounded processing time per frame, while dealing with a fixed amount of available memory. This system works under the Assumption of Insufficient Knowledge and Resource and is hence operating in real-time and open to new information which can arrive at any time, as systems such as NARS and AERA ideally also demand for perception.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
This is also referred to lifelong or incremental learning in Machine Learning [14].
- 2.
Will be extended to a more complicated kernel and in multiple directions.
- 3.
source code and the neural network models, which are trained using the Backpropagation algorithm, can be found at https://github.com/PtrMan/23R.
- 4.
References
Arel, I., Rose, D., Coop, R.: Destin: a scalable deep learning architecture with application to high-dimensional robust pattern recognition. In: 2009 AAAI Fall Symposium Series (2009)
Bromley, J., Guyon, I., LeCun, Y., Säckinger, E., Shah, R.: Signature verification using a “siamese” time delay neural network. In: Advances in Neural Information Processing Systems, vol. 6 (1993)
Carpenter, G.A., Grossberg, S.: Art 2: self-organization of stable category recognition codes for analog input patterns. Appl. Opt. 26(23), 4919–30 (1987)
Fukushima, K.: Visual feature extraction by a multilayered network of analog threshold elements. IEEE Trans. Syst. Sci. Cybernet. 5(4), 322–333 (1969)
Goertzel, B.: Perception processing for general intelligence: bridging the symbolic/subsymbolic gap. In: Bach, J., Goertzel, B., Iklé, M. (eds.) AGI 2012. LNCS (LNAI), vol. 7716, pp. 79–88. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-35506-6_9
Hammer, P., Lofthouse, T.: ‘OpenNARS for applications’: architecture and control. In: Goertzel, B., Panov, A.I., Potapov, A., Yampolskiy, R. (eds.) AGI 2020. LNCS (LNAI), vol. 12177, pp. 193–204. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-52152-3_20
Hammer, P., Lofthouse, T., Wang, P.: The OpenNARS implementation of the non-axiomatic reasoning system. In: Steunebrink, B., Wang, P., Goertzel, B. (eds.) AGI -2016. LNCS (LNAI), vol. 9782, pp. 160–170. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-41649-6_16
Koch, G.R.: Siamese neural networks for one-shot image recognition (2015)
Lofthouse, T.: ALANN: an event driven control mechanism for a non-axiomatic reasoning system (NARS). In: NARS2019 workshop at AGI (2019)
Nivel, E., et al.: Autocatalytic endogenous reflective architecture (2013)
Radford, A., et al.: Learning transferable visual models from natural language supervision (2021)
Ramachandra, B., Jones, M.: Street scene: a new dataset and evaluation protocol for video anomaly detection. In: 2020 IEEE Winter Conference on Applications of Computer Vision (WACV), pp. 2558–2567 (2019)
Robbins, H., Monro, S.: A stochastic approximation method. Ann. Math. stat., 400–407 (1951)
Silver, D.L., Yang, Q., Li, L.: Lifelong machine learning systems: beyond learning algorithms. In: AAAI Spring Symposium: Lifelong Machine Learning (2013)
Wang, C.Y., Bochkovskiy, A., Liao, H.Y.M.: Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. arXiv preprint arXiv:2207.02696 (2022)
Wang, P.: Experience-grounded semantics: a theory for intelligent systems. Cogn. Syst. Res. 6(4), 282–302 (2005)
Wang, P.: Insufficient knowledge and resources-a biological constraint and its functional implications. In: AAAI Fall Symposium: Biologically Inspired Cognitive Architectures (2009)
Wang, P., et al.: Omni-DETR: omni-supervised object detection with transformers. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 9367–9376 (2022)
Wang, P., Hammer, P., Isaev, P., Li, X.: The conceptual design of openNARS 3.1. Technical report, Temple University, Philadelphia, United States (2020)
Wang, P., Hofstadter, D.: A logic of categorization. J. Exp. Theor. Artif. Intell. 18(2), 193–213 (2006)
Wang, P., Li, X., Hammer, P.: Self in NARS, an AGI system. Front. Robot. AI 5, 20 (2018)
Wünsche, R.: A vision prototype for openNARS
Yu, J., Wang, Z., Vasudevan, V., Yeung, L., Seyedhosseini, M., Wu, Y.: Coca: contrastive captioners are image-text foundation models. arXiv preprint arXiv:2205.01917 (2022)
Acknowledgements
Special thanks to Patrick Hammer, Tony Lofthouse and Robert Johansson for valuable discussions.
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
Wünsche, R. (2023). An Adaptive Vision Architecture for AGI Systems. In: Hammer, P., Alirezaie, M., Strannegård, C. (eds) Artificial General Intelligence. AGI 2023. Lecture Notes in Computer Science(), vol 13921. Springer, Cham. https://doi.org/10.1007/978-3-031-33469-6_34
Download citation
DOI: https://doi.org/10.1007/978-3-031-33469-6_34
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-33468-9
Online ISBN: 978-3-031-33469-6
eBook Packages: Computer ScienceComputer Science (R0)