Multimodal Conversational Helpbot To Support Robot Assembly Task
Published:
Incorporate Multi-modal Context for Improving User Intent Classification Work.
- Incorporated visual feature into question answering model; improved 24% in accuracy and 0.17 in MRR compared to text-only method.
- Presented a workflow able to generalize to other domains and be adopted to smart manufacturing field.
- Submitted to ICASSP 2020
