刘启凡，南方科技大学电子与电气工程系博士后，主要从事人工智能方面研究。

研究方向

• 机器学习
• 计算机视觉
• 小样本学习
• 大模型

论文

• Liu Q, Zhao Z, Cao W, et al. Residual proportion multilayer perceptron for few-shot classification[C]//2021 IEEE International Conference on Multimedia & Expo Workshops (ICMEW). IEEE, 2021: 1-6.
• Liu Q, Cao W. Geometric algebra graph neural network for cross-domain few-shot classification[J]. Applied Intelligence, 2022, 52(11): 12422-12435.
• Zhao Z, Liu Q, Cao W, et al. Self-guided information for few-shot classification[J]. Pattern Recognition, 2022, 131: 108880.
• Liu Q, Chen Y, Cao W. Dual-domain reciprocal learning design for few-shot image classification[J]. Neural Computing and Applications, 2023: 1-14.
• Liu Q, Cao W, He Z. Cycle Optimization Metric Learning for Few-Shot Classification[J]. Pattern Recognition, 2023: 109468.
• Zhang R, Liu Q. Learning with few samples in deep learning for image classification, a mini-review[J]. Frontiers in Computational Neuroscience, 2023.

专利

*车辆检测方法、装置及计算机可读存储介质，何志权，刘启凡，曹文明，专利号：ZL 2019 1 0029834.8 *基于小波变换的渐进式深度卷积网络图像识别方法及装置，何志权，曹文明，刘启凡，专利号：ZL 2019 1 0783600.2 *图像数据的压缩传输方法、系统和计算机可读存储介质，何志权，曹文明，刘启凡，专利号：ZL 2019 1 0811971.7 *一种基于深度神经网络的OCR识别方法及装置，曹文明，刘启凡，何志权，专利号：ZL 2019 1 0904514.2