DMCL: distillation multiple choice learning for multimodal action recognition

Files
1912.10982.pdf(544.07 KB)
Accepted manuscript
Date
2019
DOI
Authors
Bargal, Sarah Adel
Garcia, Nuno
Ablavsky, Vitaly
Morerio, Pietro
Murino, Vittorio
Sclaroff, Stan
Version
Accepted manuscript
OA Version
Citation
Sarah Bargal, Nuno Garcia, Vitaly Ablavsky, Pietro Morerio, Vittorio Murino, Stan Sclaroff. "DMCL: Distillation Multiple Choice Learning for Multimodal Action Recognition." IEEE Winter Conference on Applications of Computer Vision (WACV),
Abstract
In this work, we address the problem of learning an ensemble of specialist networks using multimodal data, while considering the realistic and challenging scenario of possible missing modalities at test time. Our goal is to leverage the complementary information of multiple modalities to the benefit of the ensemble and each individual network. We introduce a novel Distillation Multiple Choice Learning framework for multimodal data, where different modality networks learn in a cooperative setting from scratch, strengthening one another. The modality networks learned using our method achieve significantly higher accuracy than if trained separately, due to the guidance of other modalities. We evaluate this approach on three video action recognition benchmark datasets. We obtain state-of-the-art results in comparison to other approaches that work with missing modalities at test time.
Description
License