Suggestions
Gadi Singer
Innovative leader in technology with extensive experience at Intel Corporation
Gadi Singer is a prominent figure in the field of artificial intelligence, currently serving as the Vice President of Intel Labs and the Director of Emergent AI Research at Intel Corporation. He has held this position since April 2020, where he leads efforts to innovate in machine intelligence and cognition. His work focuses on integrating deep learning with knowledge structures and symbolic reasoning to enhance the efficiency and explainability of AI systems.1
Professional Background
Singer's career at Intel spans several key roles:
-
Vice President, AI Products Group (June 2019 - April 2020): He was responsible for advancing inference technologies and overseeing the Intel® Nervana™ Neural Network Processor line, which won the Intel Achievement Award for innovation.1
-
Vice President, Architecture General Manager (June 2017 - May 2019): In this role, he was instrumental in driving deep learning architecture acceleration across Intel and establishing a community of AI practitioners within the company.1
-
Various Leadership Roles: Over the years, Singer has held multiple leadership positions within Intel, including roles focused on low-power technologies and cognitive computing programs, contributing significantly to the development of technologies that have shaped modern computing.1
Education and Influence
Gadi Singer is an alumnus of the Technion - Israel Institute of Technology, where he gained foundational knowledge that has supported his extensive career in technology and AI. He is recognized as a thought leader and innovator in the AI domain, frequently sharing insights on the future of AI and its applications in various industries.1
Recent Initiatives
Recently, Singer has been involved in projects aimed at making 3D content creation more accessible through technologies like Neural Object Cloning, which allows users to create detailed 3D models using smartphones. This initiative reflects his commitment to democratizing technology and enhancing creative processes in fields such as gaming and visualization.2
Overall, Gadi Singer's contributions to Intel and the broader AI landscape underscore his role as a key influencer in the ongoing evolution of artificial intelligence technologies.
Highlights
Drawing insights from biological signal processing, neuromorphic computing promises a substantially lower power solution to improve energy efficiency of visual odometry (VO) in robotics.
Published in Nature Machine Intelligence, this novel approach develops a VO algorithm built from neuromorphic building blocks called resonator networks. Demonstrated on Intel’s Loihi neuromorphic chip, the network generates and stores a working memory of the visual environment, while at the same time estimating the changing location and orientation of the camera. The system outperforms deep learning approaches on standard VO benchmarks in both precision and efficiency – relying on less than 100,000 neurons without any training.
This work is a key step in using neuromorphic computing hardware for fast and power-efficient VO and the related task of simultaneous localization and mapping (SLAM), enabling robots to navigate reliably. A companion paper explores how the neuromorphic resonator network can be applied to visual scene understanding. By formulating the generative model based on vector symbolic architectures (VSA), a scene can be described as a sum of vector products, which can then be efficiently factorized by a resonator network to infer objects and their poses.
The work demonstrates a new path for solving problems of perception and many other complex inference problems using energy efficient neuromorphic algorithms and Intel hardware.
Congratulations to researchers from the Institute of Neuroinformatics, University of Zurich and ETH Zurich, Accenture Labs, Redwood Center for Theoretical Neuroscience at UC Berkeley, and Intel Labs. Learn more about neuromorphic VO: https://lnkd.in/gJCVVMCz Learn how the VSA framework was developed for neuromorphic visual scene understanding based on a generative model (companion paper): https://lnkd.in/gjAENfpp #iamintel #Neuromorphic #Robotics
GenAI is still in the early stages of accurately generating panoramic scenes from abstract descriptions in text, a simple hand drawing, or partial views from a single image. Solving this presents great opportunities for a wide range of applications in design, education, filming, gaming, and simulation.
To address these challenges, researchers at Intel Labs introduced Language Model Assisted Generation of Images with Coherence, a generalizable AI framework that can create panorama, immersive videos, and 3D from multiple input modalities.
The AI framework uses pre-trained LLMs to control 2D diffusion models, allowing the generation of panoramic scenes with diverse 360-degree layouts, different types of scenes (indoor, outdoor, and even under water), and input modalities without fine-tuning.
The model quantitatively beats existing methods for both image- and text-to-panorama generation in both human and algorithm evaluations. Qualitatively, the model can smoothly close the 360-degree loop and generate scenes with coherent 360-degree layouts without redundant objects.
The AI framework is optimized to run on both industry AI stacks and Intel hardware, including the Intel Gaudi AI accelerator platform.
The framework has been accepted at multiple conferences, including CVPR 2024 as a highlight paper and live demonstration on the Gaudi accelerator, in addition to a demo at the ISC High Performance 2024 conference.
Read the blog: https://lnkd.in/dEBniFy9 See the code, paper, and video presentation: https://lnkd.in/dxQtFDAa
#iamintel #ArtificialIntelligence #AI #GenerativeAI #GenAI #LLMs #ComputerVision #CVPR2024
