![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/8902ee6d-4dcc-4c88-9262-a7e68bf80621.png)
在本期回顾中,我们将展示学院师生的部分科研成果(一)。这些成果凝聚着信息学院师生们的智慧与心血,是他们勇敢探索未知、自由思考问题、独立创新实践的结晶,承载着他们对卓越科研的不懈追求和坚定信念。我们期待这些成果能够激发更多灵感的火花,引发更深入、更广泛的学术讨论与交流。
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/7db76436-c6f4-4316-b478-6786a545a541.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
预测合成致死抗癌药物靶点的 机器学习方法 的基准测试
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/21cc21ef-8049-4a0d-9104-5bb0f0a608dc.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于视觉大模型的无标签场景理解
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/360c210b-982b-49b6-9875-3ce98c7cfc6a.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于无监督多色阶神经表示的 CT金属伪影 去除算法研究
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/32604358-2147-4409-b1c4-d7d59d83d0c6.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
让语言模型进行多模态推理的 “各司其职” 思维链提示
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/20fbfc51-8db1-4fe4-b858-ca33e65f9c28.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于亲和邻域优化的深度聚类
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/ef825a1d-9531-4754-9b90-070ec3a1c53d.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
用于分割中分布外检测的 异常意识测试阶段 自适应方法
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/27abd7da-9710-4f18-ae3c-b017a393bdc3.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
Q 加权变分策略优化:
基于扩散模型的强化学习
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/6a98215b-a7d4-47df-a143-8359db5d91ba.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
深度储备池光计算机
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/80530aed-4bb5-4ccb-bcd5-3cb4d2fdc159.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于螺旋波谱传播的快速被动空化成像 新方法
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/da4ac37a-2c2a-4450-9024-ada70a962166.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于 CMOS 工艺的高压恒流神经刺激芯片
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/1516afb4-9f81-4b08-bc06-eb3b5f93eadc.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于 III-V 族渐变带隙 PN 结探测器的 单像素智能微型光谱仪
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/a2737c2d-f810-4e94-9936-2fdea29baf13.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
多层循环神经网络(RNN)的可重构且 高能效的加速架构
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/2d0afaa9-627b-4c52-8cf7-6fb2b5bbd513.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
高分辨率活体脑血管温度成像方法
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/b74052cc-418d-4902-a3a0-4e637ad92a81.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
柔性直流电网线路快速灵敏差动保护新方法
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/53cbba01-3871-4841-a34a-7318dfab8e6f.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
人工智能辅助电磁超构表面逆向设计方法
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/92d648d8-a292-4f75-bf9d-9c10348d1869.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于空间稀疏性的神经网络体 渲染算法的 专用硬件加速器
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/e8828019-172b-4968-91aa-655a21ab4238.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于设计规格感知与贝叶斯优化的 SoC设计空间探索框架
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/bc262ae2-0eef-4924-8714-e232debd186b.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于大语言模型的版图热点检测方法
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/502dde0d-b221-4481-902e-48cc9418f99b.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
一种基于模拟分叉算法的 高性能 随机计算伊 辛机实现
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/0f07f7d3-8131-4610-a71a-b0845a719b62.png)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/4310e824-eef0-49ac-a295-74fda981e148.png)
基于eDRAM的低温存内计算加速器(eCIMC)
![](/_upload/article/images/32/af/9df859c145cfa5a14f8b714b3e8f/e443e231-d2ed-438e-a11f-1a9ad37007f5.png)