杨超然
博士,研究员,研究组组长
Email: yangchaoran@@gdiist.cn
主要研究方向:
类脑计算模型与系统、机器人视觉感知与控制、视觉-语言模型
目前致力于开发基于类脑计算及深度学习的开放环境视觉智能机器人算法-软件-芯片优化系统及关键技术。
个人简介:
2016年毕业于澳门大学电机与电子工程系、模拟与混合信号超大规模集成电路国家重点实验室,从事超低功耗CMOS心电图检测算法及专用数字处理器芯片的研究,获哲学博士学位。曾任澳门大学模拟与混合信号超大规模集成电路国家重点实验室技术员、职能主管。2017-2023年在深圳市华为技术有限公司从事昇腾AI处理器、深度学习算法与系统、图像与视频检测分析、数字人与具身智能、自动驾驶鸟瞰环视视觉感知等研究、开发工作,并负责多项与知名高校的技术研究合作。
曾获ISSCC参与赞助奖、第三届华为公司十大发明奖等奖项。发表SCI期刊及国际会议论文20篇,中国专利3项,美国、欧洲专利各1项。相关研究成果在IEEE Transactions on Biomedical Circuits and Systems, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, International Solid-State Circuits Conference (ISSCC) SRP, IEEE International Conference on Consumer Electronics (ICCE), Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)等期刊与会议上发表。其中发表于“IEEE Transactions on Biomedical Circuits and Systems”的关于“超低功耗心电图 QRS 波检测处理器架构”的学术 论文被两篇Nature Scientific期刊论文、一本学术专著书籍,以及该子领域论文广泛引用。
研究组简介:
自然界中生物的智能存在于其身体之中,而身体则存在于三维物理世界中。具身智能研究的动机是将智能置于三维物理世界的真实约束中,关注智能体的看、听、说、推理和行动,以更“类人”的方式对环境进行感知、理解、交互,并获取知识。具身智能这一场景约束具有普遍性,同时使得智能的研究更具针对性。具身智能领域吸引来自计算机视觉、语言、图形和机器人等跨学科的研究人员,包括李飞飞等知名研究者,解决具身智能领域的共同问题,推进数字人与机器人、自动驾驶、人机交互等多个领域的发展。近期,语言基础模型的发展为基于自然语言模型的推理能力带来了显著提升,为视觉-语言模型提供了新机会。课题组期望结合视觉感知和语言推理能力,以应对现实世界中复杂多样的场景挑战。
类脑计算是一种创新的计算技术,它借鉴了脑科学的基本原理,从算法层面模拟人脑神经元和突触的信息处理机制,并在芯片架构上突破传统“冯诺依曼”架构的限制,被普遍认为具有高能源效率的显著特点。近期,类脑计算取得了进一步的发展。如IBM继True North类脑AI芯片后,在2023年第四季发布了North Pole芯片,其在ResNet-50模型上测试的能效是相同工艺GPU竞品的25倍; Intel推出了Lolhi 2类脑芯片;OpenAI对类脑芯片进行了投资;Science Robotics期刊发表多篇类脑计算与仿生机器人的文章;世界多国以及我国的脑计划也大力支持新型类脑芯片及计算系统的研发。
具身智能机器人在开放环境中的应用需要解决三维空间中复杂物体的感知与操作问题,以及如何在本地实现高性能、高能效智能计算的问题。因此,本课题组目前专注于类脑视觉算法与系统、基于视觉-语言模型的具身智能机器人的技术研究,同时也致力于相关软硬件优化系统的研发。期望通过视觉-语言模型解决具身智能机器人的多样复杂场景适应问题,以及通过类脑计算与软硬件优化系统解决机器人本地算力的高效计算问题。
本课题组招聘博士后、工程师、实习生(软硬件系统、FPGA/ASIC、AI Processor、视觉计算、机器人学习、类脑算法),感兴趣的同学,请邮件联系。
2024级秋季入学澳门大学联合培养博士招生(已截止)
2024级秋季入学澳门理工大学联合培养博士招生(已截止)
2025级澳门大学、澳门理工大学、澳门科技大学联合培养博士招生
本课题组拟招收2025级秋季入学联合培养博士生3名,拟培养方向为面向人工智能领域具身智能、三维机器人视觉感知与控制、视觉-语言模型、类脑计算等的算法与系统方向,将根据课题组发展需求、学生兴趣与特长、学术前沿选定具体研究课题。申请人必须满足所在大学对博士课程的入学要求,在学期间享受所在大学的研究生身份,并在毕业后获得该校的博士学位。在广东省智能院开展研究期间,本课题组提供研究津贴、免费住宿、研究办公基础条件,具体以智能院及所在大学的相关规定为准。有意者请提前联系。
院招生公告:https://www.gdiist.cn/zhaos/detail/80/121
报名截止日期:
第一批为2024年10月20日;
第二批为2024年11月20日;
依据经验面试名单会提前一天收集,靠近截止日期才交的申请会大受影响。
课题组要求:
1、具有计算机、人工智能、机器人、视觉计算、电机与电脑工程、数学等相关理工科专业硕士或学士学位,学习成绩优秀;
2、具有良好的个人品德,乐于接受新事物,乐于交流,具有团队合作精神;
3、对课题组研究有兴趣,对科研有热情,有自驱力、做事有毅力;
4、有良好的Python/C++编程能力,有方向相关研究基础;
5、申请澳门大学要求985或前300高校本硕毕业;
6、具有相关顶会顶刊论文发表或高级别竞赛获奖者优先考虑;
7、在本课题组有实习经验者优先。
课题组优势:
1. 重视学生发展,关心学生成长;
2. 课题组长有工业界经验,了解学术界及业界需求,对相关领域有动手经验;
3. 关心学术界、业界前沿进展,与境内外优秀课题组及工业界有一定的交流学习机会;
4. 所处地点环境优美、舒适宜居,邻近穗深港澳,交通便利,便于交流。
代表论著:
International Journal Papers
[1] Boyu Zheng, et.al. "A Scheme for Solving Fuzzy Motion Planning for Manipulators", Submitted to IEEE Transactions on Systems, Man, and Cybernetics, 2024.
[2] Boyu Zheng, et.al. "A Hybrid-Gain ZNN and Its Application in Robotic Arm", Submitted to IEEE Transactions on Neural Networks and Learning Systems, 2024.
[3] Xie, Xinyu, et al. "Fusionmamba: Dynamic feature enhancement for multimodal image fusion with mamba." arXiv preprint arXiv:2404.09498, 2024.
[4] L. Zhao, R. Xu, T. Wang, T. Tian, X. Wang, W. Wu, C. Ieong, X. Jin, "BaPipe: Balanced Pipeline Parallelism for DNN Training", Parallel Processing Letters, 2022.
[5] Chio-In Ieong, Mingzhong Li, Man-Kay Law, Pui-In Mak, Mang I Vai, and Rui P. Martins, “A 0.45-V 147-to-375 nW ECG Compression Processor with Wavelet Shrinkage and Adaptive Temporal Decimation Architectures”, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Apr. 2017.
[6] Ming-Zhong Li, Chio-In Ieong, Man-Kay Law, Pui-In Mak, Mang-I Vai, Sio-Hang Pun, and Rui P. Martins, "Energy Optimized Subthreshold VLSI Logic Family With Unbalanced Pull-Up/Down Network and Inverse Narrow-Width Techniques," IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol. 23, no. 12, pp. 3119-3123, Dec. 2015.
[7] Chio-In Ieong, Pui-In Mak, Chi-Pang Lam, Cheng Dong, Mang-I Vai, Peng-Un Mak, Sio-Hang Pun, Feng Wan, and Rui P. Martins, "A 0.83-μW QRS Detection Processor Using Quadratic Spline Wavelet Transform for Wireless ECG Acquisition in 0.35-μm CMOS," IEEE Transactions on Biomedical Circuits and Systems, vol. 6, pp. 586-595, Dec. 2012. [Leading work in its field. Citations (Google Scholar)= 137 including citations from a book and two Nature Scientific Reports.]
International Conference Papers
[8] Hang Yu, et. al., “SDP: Spiking Diffusion Policy for Robotic Arm Action Generation”, China Biomedical Engineering Conference, oral presentation, accepted, 2024.
[9] “SDP: Spiking Diffusion Policy for Robotic Manipulation with Learnable Channel-wise Membrane Thresholds”, Nature Conference on Neuromorphic Computing, 2024, accepted.
[10] Chio-In Ieong, Mingzhong Li, Man-Kay Law, Pui-In Mak, Mang I Vai and Rui P. Martins, “A 0.45V 147-to-375nW Hardware-Efficient Real-Time ECG Processor with Lossless-to-Lossy Data Compression for Wireless Healthcare Wearables”, Presentation at the 2016 International Solid-State Circuits Conference (ISSCC 2016) Student Research Preview session, Jan. 2016, San Francisco, United States.
[11] Chio-In Ieong, Pui-In Mak, Mang-I Vai and Rui P. Martins, “Sub-μW QRS Detection Processor Using Quadratic Spline Wavelet Transform and Maxima Modulus Pair Recognition for Power-Efficient Wireless Arrhythmia Monitoring”, in Proc. of The 21st Asia and South Pacific Design Automation Conference - University Design Contest (ASP-DAC UDC 2016), Jan. 2016, Macau, China.
[12] Chio-In Ieong, Mingzhong Li, Man-Kay Law, Pui-In Mak, Mang I Vai, Peng-Un Mak, Feng Wan, Rui P. Martins, "Standard cell library design with voltage scaling and transistor sizing for ultra-low-power biomedical applications," in Proc. of 2013 IEEE International Conference of Electron Devices and Solid-State Circuits (EDSSC), Jun. 2013, Hong Kong, China
[13] Ming-zhong Li, Chio-In Ieong, Man-Kay Law, Pui-In Mak, Mang I Vai, and Rui P. Martins, "Sub-threshold standard cell library design for ultra-low power biomedical applications," in Proc. of 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 1454-1457, Jul. 2013, Osaka, Japan
[14] Chio-In Ieong, C. Dong, W. Nan, A. Rosa, R. Guimarães, M.-I Vai, P. U. Mak, “A snoring classifier based on Heart Rate Variability analysis”, Computing in Cardiology, pp. 345-348, Sep. 2011.
[15] Chio-In Ieong, Mang I Vai, Peng-Un Mak, and Pui-In Mak, "ECG heart beat detection via Mathematical Morphology and Quadratic Spline wavelet transform," in Proc. of 2011 IEEE International Conference on Consumer Electronics (ICCE), 2011, pp. 609-610, Jan. 2011, Las Vegas, USA
[16] Chio In Ieong, Mang I Vai, and Peng Un Mak, “ECG QRS Complex Detection with Programmable Hardware”, in Proc. of The 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2008). 2008: Vancouver, Canada.
[17] Chio In Ieong, Feng Wan, Mang I Vai, Peng Un Mak, “QRS Complex Detector Using Artificial Neural Network”, in Proc. of the 4th Regional Inter-University Postgraduate Electrical and Electronics Engineering Conference (RIUPEEEC2006). 2006: Macau SAR., China. [My 1st paper, NN for detection]
Patents
[18] Zhenjiang Dong, Chio-In Ieong, Hu Liu, Hai Chen, "Matrix processing method and device and logic circuit", US11734386B2, CN111010883B
[19] Zhenjiang Dong, Chio-In Ieong, Hu Liu, Hai Chen, "Convolutional Method and Device for Neural Network", CN112106034A
[20] Chenglin Zheng, Chio-In Ieong, Hai Chen, "Image processing method and apparatus based on convolutional neural network", CN110321996B
公众号