The Workshop on Open Domain Action Recognition (ODAR) will be held in conjunction with the CVPR 2017 conference.
Action recognition has received increasing attention from the computer vision and machine learning communities in the last decade. To enable the study of this problem, there exist a vast number of action datasets, which are recorded under controlled laboratory settings, real-world surveillance environments, or crawled from the Internet. Apart from the "in-the-wild" dataset, the training and test split of conventional dataset often resemble similar environments conditions, which leads to close to perfect performance on constrained dataset. This workshop aims to considered action recognition with the open domain constraint (i.e., the deployment environment are likely not to be seen in the training data space). To achieve this, we compose a new Open Domain Action Recognition Dataset, which comprised of video samples from existing dataset. Together, we carefully design an evaluation protocol, where the training, validation, and test set consists of action samples from 11 known and 1 unknown classes with similar or different camera deployment environemnt.
This workshop will consists of two tracks: (1) The challenge track will focus on the evaluation on the new datasets; (2) The regular track welcome paper that addresses open domain classification problem on related topics.
This workshop will bring together computer vision experts from academia, industry, and government who have made progress in developing computer vision tools for action recognition analysis. This workshop provides a comprehensive forum on this topic and foster in-depth discussion of technical and future research direction. It will also serve as an introduction to researchers and students curious about this field.