From Lifestyle Vlogs to Everyday Interactions
This work addresses the data bottleneck for researchers in computer vision and AI studying everyday human interactions, though it is incremental in its approach to data collection.
The paper tackled the problem of limited training data for understanding basic human interactions by using Internet Lifestyle Vlogs as a source to collect and annotate interaction-rich video data, achieving greater scale and diversity compared to explicit data gathering methods.
A major stumbling block to progress in understanding basic human interactions, such as getting out of bed or opening a refrigerator, is lack of good training data. Most past efforts have gathered this data explicitly: starting with a laundry list of action labels, and then querying search engines for videos tagged with each label. In this work, we do the reverse and search implicitly: we start with a large collection of interaction-rich video data and then annotate and analyze it. We use Internet Lifestyle Vlogs as the source of surprisingly large and diverse interaction data. We show that by collecting the data first, we are able to achieve greater scale and far greater diversity in terms of actions and actors. Additionally, our data exposes biases built into common explicitly gathered data. We make sense of our data by analyzing the central component of interaction -- hands. We benchmark two tasks: identifying semantic object contact at the video level and non-semantic contact state at the frame level. We additionally demonstrate future prediction of hands.