2016

Tubelets: Unsupervised action proposals from spatiotemporal super-voxels [PDF]
Mihir Jain, Jan van Gemert, Hervé Jegou, Patrick Bouthemy and Cees Snoek
arXiv, 2016 (under submission to IJCV)


VideoLSTM Convolves, Attends and Flows for Action Recognition [PDF]
Zhenyang Li, Efstratios Gavves, Mihir Jain and Cees Snoek
arXiv, 2016


Improved Motion Description for Action Classification [Web-link]
Mihir Jain, Hervé Jegou and Patrick Bouthemy
Frontiers in ICT, 2016

2015

Objects2action: Classifying and localizing actions without any video example [PDF]
Mihir Jain, Jan van Gemert Thomas Mensink and Cees Snoek
International Conference on Computer Vision (ICCV), Santiago, Chile, December 2015


APT: Action localization Proposals from dense Trajectories [PDF] [Code] [Precomputed tubes]
Jan van Gemert, Mihir Jain, Ella Gati and Cees Snoek
British Machine Vision Conference (BMVC), Swansea, UK, September 2015


What do 15,000 Object Categories Tell Us About Classifying and Localizing Actions? [PDF]
Mihir Jain, Jan van Gemert and Cees Snoek
Computer Vision and Pattern Recognition (CVPR), Bostom, MA, USA, June 2015
Features and other details are available on the project page

2014

Action localization by tubelets from motion [PDF] [Code]
Mihir Jain, Jan van Gemert, Hervé Jegou, Patrick Bouthemy and Cees Snoek
Computer Vision and Pattern Recognition (CVPR), Columbus, Ohio, USA, June 2014

2013

Better motion for better action recognition [PDF]
Mihir Jain, Hervé Jegou and Patrick Bouthemy
Computer Vision and Pattern Recognition (CVPR), Portland, Oregon, USA, June 2013
Code and some details of the method are available on the project page

2012

Hamming Embedding Similarity-based Image Classification [PDF]
Mihir Jain, Rachid Benmokhtar, Patrick Gros and Hervé Jegou
ACM International Conference on Multimedia Retrieval (ICMR), Hong Kong, June 2012

2011

Asymmetric Hamming Embedding [PDF]
Mihir Jain, Hervé Jegou and Patrick Gros
ACM Multimedia, Scottsdale, Arizona, USA, Nov-Dec 2011