报告题目：The development and future perspectives of Al models from our own practices

报告地点：信息楼535报告厅

报告时间：2023年05月04日（周四）上午9点30分

报告人：方辉

个人简介:方辉博士现为大连大学软件学院教授，英国拉夫堡大学计算机学院副教授，2006年在英国布拉德福德大学获得博士学位。之后于多所英国大学从事博士后研究工作及教职。主要研究领域为对比学习、迁移学习、多模态融合，多媒体信号处理和可视化分析等，在IEEE Conference on Computer Vision and Pattern Recognition (CVPR，CCF A类会议)，AAAI Conference on Artificial Intelligence (AAAI，CCF A类会议)，IEEE Transactions on Visualization and Computer Graphics （CCF A类，中科院1区，影响因子5.226），Pattern Recognition （CCF B类，中科院1区，影响因子8.518），IEEE Transactions on Multimedia（CCF B类，中科院1区，影响因子8.182），IEEE Transactions on Circuits and Systems for Video Technology （CCF B类，中科院1区，影响因子5.859）等国际知名学术期刊及会议上发表60余篇论文。主持并参与国家级项目多项，具有充分的项目研究经验。

摘要：Advanced AI techniques have been developed rapidly in recent years. Many deep neural network models have been successfully applied in real-world applications. However, several key challenges hinder the further use of these new models more widely. In this talk, after identifying these challenges, our possible solutions are discussed to shed light on future perspectives of AI techniques. I will present our recent papers, including weakly supervised learning, attention-based models, contrastive learning, few shot based learning, spiking neuron network and their applications.