Search

Teaching Reform and Practical Exploration of Courses Based on Deep Learning for Small Target Detection

Author(s)

Yeqiang Zheng1, Haiwei Tan2, Zhiqiang Zhou2

Affiliation(s)

¹Publicity Department, Yulin Normal University, Yulin, Guangxi, China ²School of Artificial Intelligence, Yulin Normal University, Yulin, Guangxi, China

Abstract

Faced with the AI industry’s urgent demand for compound talents in deep learning-based small target detection and prominent university teaching pain points like the theory-practice disconnection, lagging content, flawed practical system and inadequate innovative ability cultivation, this paper designs a systematic curriculum reform plan centering on the deep learning-based small target detection method, based on constructivism and project-oriented theories. Through investigating 5 universities and industrial talent demands, the reform is implemented in four dimensions: teaching content, methods, practical system and evaluation mechanism, with 86 AI majors as pilots. It reconstructs a three-tier content system of "basic theory-core technology-engineering application", adopts a diversified teaching method integrating topic-driven and blended learning, builds a four-tier practical system and a trinity evaluation system. The reform significantly improves students’ academic performance, engineering and research abilities, exercises teachers’ competencies, forms a "theory-practice-research" teaching mode and establishes a school-enterprise education mechanism. It effectively addresses traditional teaching problems, connects talent training with industrial demands, and provides a replicable plan for similar course reforms.

Keywords

Deep Learning; Small Target Detection; Teaching Reform; Practical Teaching; Topic-Driven; Talent Training.

References

[1] Li F F, Karpathy A, Jia Y Q. Computer Vision: Algorithms and Applications. Beijing: China Machine Press, 2021. [2] Chen Y, Li J. Research and Teaching Practice of Small Target Detection Technology Based on Anchor-Free Algorithm. Computer Engineering and Science, 2023, 45(S1): 201-206. [3] Zhang S, Li S. Reform of the Trinity Teaching Mode of "Theory-Experiment-Project" in the Target Detection Course. Computer Education, 2023(05): 123-126. [4] Liu M, Wang H. Teaching Reform of Deep Learning Course from the Perspective of Industry-University-Research-Application Integration. Chinese Vocational and Technical Education, 2023(18): 89-93. [5] Zhang L, Wang L. Application of Project-Oriented Teaching in the Target Detection Course. Software Guide, 2022, 21(10): 234-238. [6] Zhang W, Cai Y F, Ye L T, et al. Real-Time Transformer Small Target Detection Model Based on Enhanced Feature Extraction and Pyramid Structure. Computer Science, 2025, 52(S2): 375-385. [7] He K M, Zhang X Y, Ren S Q, et al. Deep Residual Learning for Image Recognition. Chinese Journal of Computers, 2020, 43(01): 1-15. [8] Ni J, Shen A, Wang Z. UAV Small Target Detection Based on Improved YOLOv7tiny. Computer Engineering and Design, 2025, 46(11): 3065-3073. [9] Zhu D Y, Lv J M, Zhou P, et al. Method for Ground Small Target Detection and Recognition of Micro UAV SAR. Journal of Nanjing University of Aeronautics and Astronautics (Natural Science Edition), 2025, 57(05): 781-798. [10]Wang W, Zhao L. Teaching Reform and Practice of Deep Learning Course for Artificial Intelligence Major. Higher Education of Sciences, 2024(01): 98-102.