Open Data Science Conference (ODSC)的动态

Thinking back to 2016, I can remember discussions about preparing students for a future shaped by automation, robotics, and the nascent field of artificial intelligence. Fast forward to today, and that once-distant future has arrived with startling velocity.The urgent question echoing throughout higher education today shouldn’t be about preparing students anymore. Now, we must ask: Is our entire university system – from staff to faculty to administration – ready for the AI revolution? We have a dual responsibility: to our students and to our employees. Just as we strive to future-proof our curricula, we must also proactively equip our staff and faculty with the skills and knowledge to work alongside AI. This isn't merely about job security; it's about preparing for the full potential of human-AI collaboration to enhance every facet of the educational experience.

A new framework has been developed for generating human motions from language prompts. Advanced machine learning models are revolutionizing content creation, with a recent breakthrough in human motion generation presented by researchers at the Beijing Institute of Technology and Peking University. Their framework, building on previous work called HUMANIZE, utilizes scene affordance to enhance language-guided human motion generation, offering improved 3D grounding and generalization capabilities. This innovative approach, showcased at The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2024, holds promise for applications in filmmaking and robotics training data generation. The team aims to address data scarcity and enhance inference efficiency for practical implementation in the future. #MachineLearning #ContentGeneration #HumanMotion #AIResearch #ComputerVision

As the semester draws to a close, I’d like to some reflections on teaching ME 686: Advanced Topics in Human-Robot Interaction. This course was an exciting exploration of AI, robotics, and ethics, designed to prepare students for cutting-edge challenges in the rapidly evolving world of technology. One highlight was our incredible guest speakers, who shared their expertise on topics ranging from social robotics for dementia care to the role of AI in business intelligence. These sessions enriched our discussions and inspired students to connect theory with practice. Another key aspect was our active learning approach, which encouraged students to engage deeply through research presentations, peer reviews, and hands-on assignments. It was a reminder of how impactful learning can be when it’s both collaborative and applied. I’m grateful to my students for their curiosity and hard work, as well as the guest speakers and colleagues who contributed to the course’s success. Looking ahead, I’m excited to refine this course further and explore new opportunities in teaching and research. Here’s to a semester of growth, learning, and collaboration! #HumanRobotInteraction #AI #Robotics #ActiveLearning #EthicsInAI

To what extent has deep reinforcement learning (DRL) seen real-world success in robotics? A new survey from the University of Texas at Austin and University of Virginia sheds light on how DRL is shaping the future of robotics-highlighting real-world success, current challenges, and exciting pathways ahead. The researchers introduced a novel taxonomy to evaluate DRL applications across four dimensions: ????????????????????????, ?????????????? ??????????????????????, ???????????????? ????????????????????, ????????-?????????? ?????????????? ????????????. The research suggests the exciting pathways lie in developing more efficient algorithms, integrating diverse skills to tackle complex tasks, and leveraging large-scale models to enhance adaptability and generalization. Congrats to the research team, Chen Tang, Ben Abbatematteo, Jiaheng Hu, Rohan Chandra, Roberto Martín-Martín, and Peter Stone! ?? Read the full paper: https://lnkd.in/gaBYSHye ?? What challenges do you think are holding back DRL in robotics? Let's discuss below! P.S. As an aspiring student researching robotics, this was very insightful to see where DRL is headed in robotics! #Robotics #AI #DRL #MachineLearning #Tech #Innovation #Engineering

#CSUSB’s xREAL Lab wraps up another amazing semester! Faculty fellows and students are innovating with XR, AI, and robotics. ?? From tackling obesity and police reform to enhancing K-12 education with AR, the lab is transforming teaching, learning, and research. Read more: https://bit.ly/40iaEaF #CoyotePride #HumanImpact

?? Curious about the future of artificial intelligence and robotics? Join Daniela Rus, a pioneer in the field, as she unveils the possibilities. In her recent TED Conferences talk, 'How AI Will Step Off the Screen and into the Real World,' Daniela sheds light on the fascinating convergence of AI and robotics, unlocking a host of possibilities for our everyday lives. Daniela introduces the concept of "liquid networks," a groundbreaking class of AI inspired by the neural processes of simple organisms. These networks promise to revolutionize the way machines process information, paving the way for what Daniela calls "physical intelligence." As the Andrew (1956) and Erna Viterbi Professor of Electrical Engineering and Computer Science, and Director of the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology, Daniela's research spans robotics, artificial intelligence, and data science. Her vision is clear: to integrate machines into our lives in a way that enhances human potential and supports us in both cognitive and physical tasks. This fascinating talk explores the transformative potential of AI and robotics, as Daniela paints a compelling picture of a future where technology enhances our lives in ways we've only dreamed of.... Video source: TED Conferences #AI #Robotics #Innovation #FutureTech #Technology

Today at #ICRA2024 IEEE RAS ICRA Dominik Joho and I are chairing the "Representation Learning I" session?from 13:30-15:00 at Annex Hall - AX-206, Yokohama, Japan. Representation learning has become its own field in the machine learning community. I am thrilled that there are dedicated sessions for robotics at #ICRA2024. Here is the list of 9 papers from institutes such as Bosch Center for Artificial Intelligence (BCAI) Carnegie Mellon University KUKA CSIRO Robotics NVIDIA Tsinghua University University of Pennsylvania ETH Zürich among many others: Tatiya, Gyan, Jonathan Francis, Ho-Hsiang Wu, Yonatan Bisk, and Jivko Sinapov. "Mosaic: Learning unified multi-sensory object property representations for robot perception." ? https://lnkd.in/dww-ViwB Ren, Hanwen, and Ahmed H. Qureshi. "Neural Rearrangement Planning for Object Retrieval from Confined Spaces Perceivable by Robot's In-hand RGB-D Sensor." ? https://lnkd.in/dHGWMPQG He, Yuze, Peng Wang, Yubin Hu, Wang Zhao, Ran Yi, Yong-Jin Liu, and Wenping Wang. "MMPI: a Flexible Radiance Field Representation by Multiple Multi-plane Images Blending."?? https://lnkd.in/dNZfrZEX Joho, Dominik, Jonas Schwinn, and Kirill Safronov. "Neural Implicit Swept Volume Models for Fast Collision Detection." https://lnkd.in/dgrarn5S Hausler, Stephen, David Hall, Sutharsan Mahendren, and Peyman Moghadam. "Reg-NF: Efficient Registration of Implicit Surfaces within Neural Fields.”, https://lnkd.in/dbmcdjua Zhou, Junsheng, Xin Wen, Baorui Ma, Yu-Shen Liu, Yue Gao, Yi Fang, and Zhizhong Han. "3D-OAE: Occlusion auto-encoders for self-supervised learning on point clouds." https://lnkd.in/dWji7D3a Shi, Junyao, Jianing Qian, Yecheng Jason Ma, and Dinesh Jayaraman. "Composing Pre-Trained Object-Centric Representations for Robotics From" What" and" Where" ? https://lnkd.in/d8ZfbEGk Kuo, Chia-Liang, Yu-Wei Chao, and Yi-Ting Chen. "SKT-Hang: Hanging Everyday Objects via Object-Agnostic Semantic Keypoint Trajectory Generation." ? https://lnkd.in/dTv3H_kR Li, Lei, Alexander Liniger, Mario Millhaeusler, Vagia Tsiminaki, Yuanyou Li, and Dengxin Dai. "Object-centric Cross-modal Feature Distillation for Event-based Object Detection." https://lnkd.in/df7qFs8j?

Robots Meet Real-World Challenges: Vision, Planning, and the Future of Automation "Know your limits" applies not just to humans but also to robots – especially when it comes to mastering complex tasks. Here's what MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) researchers Nishanth Kumar and co-lead author Aidan Curtis are doing to help robots overcome those limits, with the help of LLM. https://lnkd.in/d48x-Ez8

?? Exciting News! ?? I'm thrilled to share a fantastic resource ( downloadable) for anyone diving into the world of Computer Vision, Robotics, and Machine Learning. ?? Credit : Jean de Dieu Nyandwi Check out the comprehensive book on Linear Algebra for Computer Vision, Robotics, and Machine Learning! This book covers essential topics such as: Vector Spaces, Bases, Linear Maps: Understand the fundamentals of linear combinations, independence, and the dual space. Matrices and Linear Maps: Learn about matrix representation, composition, and the effects of changing bases. Haar Bases, Haar Wavelets, Hadamard Matrices: Explore signal compression, multiresolution analysis, and digital image transformation. Whether you're a student, researcher, or professional, this book is a must-have to deepen your understanding and enhance your skills in these cutting-edge fields. ?? Link to the book https://lnkd.in/duvjNvmh #ComputerVision #Robotics #MachineLearning #LinearAlgebra #DataScience #AI #DeepLearning #TechEducation #GenerativeAI

?? Exciting Learning Experience Completed! ?? I am thrilled to share that I have successfully completed a 3-Day Workshop on Robotics and AI! This intensive workshop provided me with a deep dive into the fascinating world of robotics and artificial intelligence, covering both theoretical foundations and practical applications. Key takeaways from the workshop include: ?? Robotics: Gained hands-on experience in designing, building, and programming robots, enhancing my understanding of robotics hardware and software integration. ?? Artificial Intelligence: Explored AI concepts, including machine learning algorithms, neural networks, and computer vision, and learned how to apply them in real-world scenarios. ?? Collaboration and Networking: Engaged with industry experts and fellow participants, fostering valuable connections and gaining diverse perspectives on the future of robotics and AI. A big thank you to the workshop organizers and instructors for delivering such a comprehensive and engaging learning experience. This workshop has significantly enriched my knowledge and skills, and I am excited to apply what I’ve learned in future projects. Looking forward to connecting with fellow robotics and AI enthusiasts and professionals! #Robotics #ArtificialIntelligence #AI #Workshop #LearningExperience #ProfessionalDevelopment #TechSkills #Innovation