Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

The University of British Columbia

$42,000.00

Flexible Diffusion Modeling for Planning and Control

EF 2024-25

Résumé
The research proposal outlines a project focused on Flexible Diffusion Modeling for Planning and Control, building on recent advancements in diffusion models. These models, capable of flexible, amortized, test-time conditioning, mark a transformative moment in AI, akin to the impact of large language models in natural language processing. The proposal details and highlights the recent success of flexibly conditioned diffusion models in generating high-dimensional, realistic video data and solving combinatorially hard problems. The specific research proposal involves exploring the capabilities of diffusion models in joint distributions over multi-modal domains such as video, speech, and action. The aim is to design and learn task distributions for training and evaluation that enable effective test-time conditioning, facilitating model predictive control and agent behavior cloning in complex, life-like domains. The ultimate goal is to make significant strides towards embodied AI, using and improving diffusion models as a powerful conditional generative AI tool. The proposal emphasizes the need for new datasets and benchmarks to evaluate and compare generative AI models in these complex, embodied AI agent domains, signaling a shift in the focus of generative AI research toward generating interesting data for training and innovating illuminating tasks.
Programme
Discovery Grants Program - Individual
Emplacement
British Columbia, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
RGPIN-2024-06371
Name nom
Wood, Frank
Department département
Computer Science
Organization id
2
Institution établissement
The University of British Columbia
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
42000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2604
Research subject group en
Robotics
Research subject group fr
Robotique
Research subject en
Intelligent systems applications
Research subject fr
Applications de systèmes intelligents
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Flexible Diffusion Modeling for Planning and Control
Keywords
artificial intelligence; machine learning; generative modeling; diffusion models; bayesian modeling; inference; amortized inference; embodied artificial intelligence; reinforcement learning; graphical models
Application summary
The research proposal outlines a project focused on Flexible Diffusion Modeling for Planning and Control, building on recent advancements in diffusion models. These models, capable of flexible, amortized, test-time conditioning, mark a transformative moment in AI, akin to the impact of large language models in natural language processing. The proposal details and highlights the recent success of flexibly conditioned diffusion models in generating high-dimensional, realistic video data and solving combinatorially hard problems. The specific research proposal involves exploring the capabilities of diffusion models in joint distributions over multi-modal domains such as video, speech, and action. The aim is to design and learn task distributions for training and evaluation that enable effective test-time conditioning, facilitating model predictive control and agent behavior cloning in complex, life-like domains. The ultimate goal is to make significant strides towards embodied AI, using and improving diffusion models as a powerful conditional generative AI tool. The proposal emphasizes the need for new datasets and benchmarks to evaluate and compare generative AI models in these complex, embodied AI agent domains, signaling a shift in the focus of generative AI research toward generating interesting data for training and innovating illuminating tasks.