WILLOW - 2016 - Annual activity report

WILLOW

WILLOW - 2016

Project-Team Willow

Members

Overall Objectives

Statement

Research Program

3D object and scene modeling, analysis, and retrieval
Category-level object and scene recognition
Image restoration, manipulation and enhancement
Human activity capture and classification

Application Domains

Introduction
Quantitative image analysis in science and humanities
Video Annotation, Interpretation, and Retrieval

Highlights of the Year

New Software and Platforms

NetVLAD: CNN architecture for weakly supervised place recognition
Unsupervised learning from narrated instruction videos
ContextLocNet: Context-aware deep network models for weakly supervised localization
Long-term Temporal Convolutions for Action Recognition

New Results

3D object and scene modeling, analysis, and retrieval
Category-level object and scene recognition
Image restoration, manipulation and enhancement
Human activity capture and classification

Bilateral Contracts and Grants with Industry

Facebook AI Research Paris: Weakly-supervised interpretation of image and video data (Inria)
Google: Learning to annotate videos from movie scripts (Inria)
Google: Structured learning from video and natural language (Inria)
MSR-Inria joint lab: Image and video mining for science and humanities (Inria)

Partnerships and Cooperations

National Initiatives
European Initiatives
International Initiatives
International Research Visitors

Dissemination

Promoting Scientific Activities
Teaching - Supervision - Juries
Popularization

Bibliography

Publications of the year

Previous |

Home | Next next

Section: New Software and Platforms

Long-term Temporal Convolutions for Action Recognition

Open source release of the software package for our paper "Long-term Temporal Convolutions for Action Recognition" [20]. It provides code for training models, testing on standard datasets and trained models. Links are available at our project page http://www.di.ens.fr/willow/research/ltc/.

Previous |

Home | Next next