Seminars home
High-level information representation in the web of things
at
Speaker: Gilles Privat, senior research scientist at Orange Labs, Meylan, France.
Abstract: The transformative novelty of the Internet of Things has, for all the hype that surrounds it, been widely misunderstood and clouded by a surfeit of ill-informed interpretations, whether blinkered or catch-all. We try to bring out its distinctive underpinnings and broader scientific relevance as a mapping between physical environments and graph-based informational representations.
The first key idea is to abstract away the constraints, protocols and identifiers originating from network connectivity, placing IoT representations on a plane where relationships between IoT “things” become independent of underlying networks (of the radio/fibre/cable kind) and their protocols of whatever layer. From this viewpoint, things need not even be connected through a network to get integrated in the Web of Things (WoT), defined as a directed graph, extending and overlaying the internet of devices (i.e. networked transducers), down to which it gets all too often demoted. This quantitatively (~trillion-nodes!) and qualitatively different Web of Things may draw inspiration from social networks, which have also abstracted away the networks which they distantly rely upon. From there upwards, the social web has become the graph par excellence, the crucible of graph-based information systems and a potential inspiration for the architecture of an overarching Web of Things platform.
Yet the Web of Things is, for all intents and purposes, a much bolder proposition than the social web, if only because “things” are so widely different in nature, capabilities and reach. Providing the interoperable representations and platforms that may support information mediation for WoT applications across the board requires a huge leap from the current situation of watertight silos and stovepipe applications that dominate the IoT landscape. Just categorizing “things” properly, across all relevant application domains, independently of their legacy identifiers, is a momentous undertaking… The original web and its expansion towards a knowledge graph and “linked open data” point in a direction whereby a proper semantics of things may emerge from the structure of the graphs they make up, as it gets integrated into the “Giant-Global-Graph” envisioned by Tim Berners-Lee.
To make robots autonomous in real-world everyday spaces, they should be able to learn from their interactions within these spaces, how to best execute tasks specified by non-expert users in a safe and reliable way. To do so requires sequential decision-making skills that combine machine learning, adaptive planning and control in uncertain environments as well as solving hard combinatorial optimisation problems. Our research combines expertise in reinforcement learning, computer vision, robotic control, sim2real transfer, large multimodal foundation models and neural combinatorial optimisation to build AI-based architectures and algorithms to improve robot autonomy and robustness when completing everyday complex tasks in constantly changing environments.
VISION
The research we conduct on expressive visual representations is applicable to visual search, object detection, image classification and the automatic extraction of 3D human poses and shapes that can be used for human behavior understanding and prediction, human-robot interaction or even avatar animation. We also extract 3D information from images that can be used for intelligent robot navigation, augmented reality and the 3D reconstruction of objects, buildings or even entire cities.
Our work covers the spectrum from unsupervised to supervised approaches, and from very deep architectures to very compact ones. We’re excited about the promise of big data to bring big performance gains to our algorithms but also passionate about the challenge of working in data-scarce and low-power scenarios.
Furthermore, we believe that a modern computer vision system needs to be able to continuously adapt itself to its environment and to improve itself via lifelong learning. Our driving goal is to use our research to deliver embodied intelligence to our users in robotics, autonomous driving, via phone cameras and any other visual means to reach people wherever they may be.
This web site uses cookies for the site search, to display videos and for aggregate site analytics.
Learn more about these cookies in our privacy notice.
Cookie settings
You may choose which kind of cookies you allow when visiting this website. Click on "Save cookie settings" to apply your choice.
This content is currently blocked. To view the content please either 'Accept social media cookies' or 'Accept all cookies'.
For more information on cookies see our privacy notice.