With the aim of building next generation virtual assistants that can handle multimodal inputs and perform multimodal actions, we introduce two new datasets (both in the virtual shopping domain), the annotation schema, the core technical tasks, and the baseline models. The code for the baselines and the datasets will be opensourced.
Other
131
stars
36
forks
source link
Fixes #8; excludes intents as attributes for SIMMC-Fashion #11
Due to some inconsistencies in the intent annotations, about 0.5% of the SpecifyInfo actions have intents in their attributes. This commit fixes those by explicitly checking for presence of attribute while extracting them by verifying the presence of a .attribute suffix.
Due to some inconsistencies in the intent annotations, about 0.5% of the
SpecifyInfoactions have intents in their attributes. This commit fixes those by explicitly checking for presence of attribute while extracting them by verifying the presence of a.attributesuffix.