Asset Bank was one of the first digital asset management solutions to offer auto-tagging capabilities. Martin discusses what we did, and learned, in an interview for Tagging.tech. He spoke to Henrik de Gyor, Chief Consultant at Another DAM Consultancy, about the potential that machine learning has for the DAM industry and the progress that needs to be made for it to be truly useful for users.
Martin discusses:
- The MVP we developed in record time to add auto-tagging functionality to Asset Bank, and to gain feedback from users on the visual recognition capabilities of API providers Clarifai and Google Cloud Vision.
- How the tags suggested by these APIs are probably not yet accurate enough to be used on their own for most digital asset management needs. However, combined with human oversight they have the potential to save a lot of time.
- Where the API vendors should focus their efforts if they want to appeal to the (huge!) DAM market.
- How a hybrid of machine-suggested tags and human input can work well to save time in the tagging process.
