本文分析了当前新一代信息技术革命为电子信息类专业人才培养模式带来的挑战,探讨了在现代产业学院背景下电子信息类专业与行业企业进行深度合作的模式和机制,以辽宁工业大学为例,介绍了学校依托“智联电子现代产业学院”培养电子信息类专业复合型人才的经验。校企协作共建产出导向的“专业课程–能力体系模型”和校企合作课程体系,形成了“全周期”、“个性化”、“长期化”的培养机制。同时,校企合作打造具有较高教学能力、科研能力、工程实践能力和职业发展能力的“双师双能型”师资队伍。建立了校、企、院三级教学质量监控体系和教学质量评价机制。通过实施上述改革措施,电子信息类专业人才培养质量显著提升,学校产学研合作成果丰硕,教师队伍结构得到优化,教学水平和工程实践能力显著提升。The challenges posed by the current revolution in new-generation information technologies to the talent cultivation model for electronic information-related majors are analyzed in this paper. The modes and mechanisms for deep collaboration between electronic information disciplines and industry enterprises under the context of modern industry colleges are explored. Taking Liaoning University of Technology as an example, the institution’s experience in cultivating interdisciplinary talents through the “Smart IoT Electronics Modern Industry College” is introduced. An output-oriented “professional curriculum - competency system model” and a collaborative industry-academia course system are collaboratively established by universities and enterprises, resulting in the formation of a “whole-cycle”, “personalized” and “long-term” training mechanism. Furthermore, a “dual-qualified and dual-ability” with strong teaching capabilities, research expertise, engineering practice skills, and career development competencies is jointly cultivated. A three-tier (u
对于非负矩阵分解的语音增强算法在不同环境噪声的鲁棒性问题,提出一种稀疏正则非负矩阵分解(SRNMF)的语音增强算法。该算法不仅考虑到数据处理时的噪声影响,而且对系数矩阵进行了稀疏约束,使其分解出的数据具有较好的语音特征。该算法首先在对语音和噪声的幅度谱先验字典矩阵学习的基础上,构建联合字典矩阵,然后更新带噪语音幅度谱在联合字典矩阵下的系数矩阵,最后重构原始纯净语音,实现语音增强。实验结果表明,在非平稳噪声和低信噪比(小于0 d B)条件下,该算法较好地削弱了噪声的变化对算法性能的影响,不仅有较高的信源失真率(SDR),提高了1~1.5个数量级,而且运算速度也有一定程度的提高,使得基于非负矩阵分解的语音增强算法更实用。
