Facebook has announced that it will open-source its AugLy Python Library, a Resource that provides for enhanced AI processing. AugLy helps AI researchers better process content the kind of augmentations that are common across the internet. Aiding in the analysis of common and impactful modifications.
The library provides more than 100 data augmentations spanning various modalities that include audio, image, video, and text. These augmentations could include anything from cropping an image, to changing the pitch of a voice recording. By providing access to augmentation data across modalities Facebook believes that AI researchers will better be able to analyze the big picture behind content. The announcement notes that:
“Combining different modalities -- such as text and images or audio and video -- using real-world augmentations can help machines better understand complex content. The meaning of the text phrase “love the way you smell today,” for example, changes entirely when overlayed on an image of a skunk.”
The company notes that they have derived these augmentations from the type of content modification that they see across the Facebook ecosystem. As such, this library is especially powerful for researchers that are working in the social media space.
Interested developers can access the data via GitHub.