IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 1 Brain-Inspired Learning, Perception, and Cognition: A Comprehensive Review Licheng Jiao , Fellow, IEEE, Mengru Ma , Pei He , Xueli Geng , Graduate Student Member, IEEE, Xu Liu , Senior Member, IEEE, Fang Liu , Senior Member, IEEE, Wenping Ma , Senior Member, IEEE, Shuyuan Yang , Senior Member, IEEE, Biao Hou , Senior Member, IEEE, and Xu Tang, Senior Member, IEEE Abstract— The progress of brain cognition and learning mech- anisms has provided new inspiration for the next generation of artiﬁcial intelligence (AI) and provided the biological basis for the establishment of new models and methods. Brain science can effectively improve the intelligence of existing models and systems. Compared with other reviews, this article provides a comprehensive review of brain-inspired deep learning algorithms for learning, perception, and cognition from microscopic, meso- scopic, macroscopic, and super-macroscopic perspectives. First, this article introduces the brain cognition mechanism. Then, it summarizes the existing studies on brain-inspired learning and modeling from the perspectives of neural structure, cognitive module, learning mechanism, and behavioral characteristics. Next, this article introduces the potential learning directions of brain-inspired learning from four aspects: perception, cognition, understanding, and decision-making. Finally, the top-ten open problems that brain-inspired learning, perception, and cognition currently face are summarized, and the next generation of AI technology has been prospected. This work intends to provide a quick overview of the research on brain-inspired AI algorithms and to motivate future research by illuminating the latest developments in brain science. Index Terms— Artiﬁcial intelligence (AI), brain-inspired algo- rithms, cognition, learning, perception. I. I NTRODUCTION T HE brain is the most complex and perfect dynamic information processor in natural evolution, which is the core of human intelligence [1]. Many important advances in artiﬁcial intelligence (AI) research reﬂect the fact that even with partial reference to brain mechanisms, it can effectively improve the intelligence level of existing models [2]. However, there is still a big gap between the current AI algorithm and the brain [3]. In order to truly approach or even surpass human-level AI, it is extremely important to conduct more in-depth research and reference on the function, structure, and processing mechanisms of the brain [4]. In the past two or three decades, the rapid progress of neuroscience and AI [5], [6] has been made in related ﬁelds. Advances in brain and neuroscience, especially with the help of new technologies and devices, support many efforts to obtain multilevel biological evidence of the brain through various experimental approaches [7], [8]. The structural and Manuscript received 1 December 2023; revised 6 May 2024; accepted 12 May 2024. (Corresponding author: Licheng Jiao.) The authors are with the Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education, International Research Center for Intelligent Perception and Computation, School of Artiﬁcial Intelligence, Hangzhou Institute of Technology, Xidian University, Xi’an 710071, China (e-mail: lchjiao@mail.xidian.edu.cn). Digital Object Identiﬁer 10.1109/TNNLS.2024.3401711 functional basis of biological intelligence can be revealed from different aspects [9]. Neuroscience and AI have a long history of intertwining. Neuroscience provides a rich source of inspiration and cognitive basis for the development of AI algorithms [10], [11]. The brain-inspired AI algorithm seeks inspiration from the biological brain, which introduces advanced results in neuroscience, cognitive science, and psychology [12], [13]. Brain-inspired algorithms can alleviate the limitations of cur- rent models and methods and improve the performance of neural computing (such as accuracy, robustness, adaptability, generalizability, and interpretability). Recently, several brain-inspired reviews have been published. Poo et al. [14] provided a general summary of the overall progress in the ﬁeld of international brain science and intelligent technology. Hassabis et al. [10] surveyed the history of the sciences of AI and neurology. In [15], it gave a thorough history of brain-inspired AI and its architecture. In [16], it concentrated on the link between brain research and improvements in computer vision. In [17], it investigated novel brain imaging techniques in order to uncover the mysteries of brain science and created brain dynamic connection maps. Schmidgall et al. [18] presented a comprehensive review of current brain-inspired learning representations in artiﬁcial neural networks. In [19], it highlighted the connection between AI and neuroscience. Ji et al. [20] studied the brain’s effective connection network in different groups of people. In [21], it provided a comprehensive survey of learning-based human–machine dialog systems. In [22], it presented the application of learning-based monocular methods to autonomous systems. Based on these literature review comparative analyses, this work aims to provide a comprehensive review and clas- siﬁcation of recent brain-inspired learning, perception, and cognition, as well as to summarize potential research directions and the top-ten open issues that need to be addressed urgently in the future. The following are the primary contributions of this article. 1) Comprehensiveness and Readability: This article pro- vides a comprehensive review of more than 200 brain neuroscience advances and brain-inspired algorithms. We summarize the existing brain-inspired learning and modeling research in terms of neural structure, cognitive modules, learning mechanisms, and behavioral charac- teristics. 2) Novelty: This article summarizes the seven characteris- tics of the brain in processing information from different perspectives: sparsity, learning, selectivity, directionality, plasticity, knowledge, and diversity. 2162-237X © 2024 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See https://www.ieee.org/publications/rights/index.html for more information. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. Authorized licensed use limited to: XIDIAN UNIVERSITY. Downloaded on July 08,2024 at 11:39:42 UTC from IEEE Xplore. Restrictions apply.