The One About Computer Vision
Computer vision is evolving from recognizing images to understanding and interacting with the world. Join us to explore how vision systems are built and deployed today alongside research insights into how machines can develop deeper visual understanding.
More About the Sharings
Cody Kamin (Director of Partnerships, Datature) will share on "Vision Language Models"
Vision-Language Models represent a significant leap forward from traditional computer vision by integrating language understanding, creating more holistic and human-like AI systems. However, fine-tuning VLMs for specific tasks presents unique challenges that go beyond standard machine learning workflows. Cody will explore three critical obstacles: managing and aligning multimodal datasets that combine visual and language components, the scarcity of experts who understand both computer vision and LLM architectures, and coordinating complex multi-GPU training processes. Learn how Datature's no-code platform addresses these challenges and discover practical approaches for successful VLM implementation in production environments. (Technical Level: 100 - 200)
Jiajun Wu (Assistant Professor, Stanford University) will share on "Understanding Visual Intelligence Through Physical Intrinsics".
Much of our visual world has intrinsic, physical structure: scenes composed of objects with their own geometry, texture, material, and physical properties. But how can we infer and represent such structure from raw visual data without hampering neural network expressiveness? Jiajun will discuss recent efforts in machine visual understanding, reconstruction, and generation, contrasting two technical approaches: leveraging intrinsics as powerful inductive biases versus grounding pre-trained vision foundation models onto intrinsics. He will demonstrate how we can now build visual intelligence that infers object shape, texture, material, and physics from a single image or video, with applications in controllable, action-conditioned 4D visual world understanding, generation, and interaction. (Technical Level: 200-300)
More About the Speakers
Cody Kamin is Director of APAC Sales & Partnerships at Datature, where he works with teams across industries to deploy Vision AI systems. He has nearly a decade of experience in autonomous systems and AI commercialization, including leadership roles at Motional and nuTonomy, where he helped launch the world’s first public autonomous ride-hailing pilot in Singapore. Cody holds an MSc from MIT and a BS in Aerospace Engineering from Georgia Tech.
Dr Jiajun Wu is an Assistant Professor of Computer Science and, by courtesy, of Psychology at Stanford University, working on computer vision, machine learning, robotics, and computational cognitive science. Before joining Stanford, he was a Visiting Faculty Researcher at Google Research. He received his PhD in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology. Wu's research has been recognized through the Young Investigator Programs (YIP) by ONR and by AFOSR, the NSF CAREER award, the Okawa research grant, the AI's 10 to Watch by IEEE Intelligent Systems, paper awards and finalists at ICCV, CVPR, SIGGRAPH Asia, ICRA, CoRL, and IROS, dissertation awards from ACM, AAAI, and MIT, the 2020 Samsung AI Researcher of the Year, and faculty research awards from Google, J.P. Morgan, Samsung, Amazon, and Meta.
More About the Series
AI Wednesdays is Lorong AI’s weekly gathering, bringing together practitioners, researchers and innovators for technical discussions on research insights, product development and engineering practices.
Get involved: Learn more about Lorong AI | Speaker Sign-up | WhatsApp Community | LinkedIn | X