Multimodal Interactions Using Pretrained Unimodal Models for SIMMC 2.0
This work addresses the problem of multimodal interaction for conversational AI systems, but it is incremental as it builds on existing pretrained models and focuses on specific benchmark tasks.
The paper tackled the challenge of multimodal conversations in the SIMMC 2.0 dataset by proposing a model that combines image and text understanding, achieving third-best performance in two subtasks and runner-up in another.
This paper presents our work on the Situated Interactive MultiModal Conversations 2.0 challenge held at Dialog State Tracking Challenge 10. SIMMC 2.0 includes 4 subtasks, and we introduce our multimodal approaches for the subtask \#1, \#2 and the generation of subtask \#4. SIMMC 2.0 dataset is a multimodal dataset containing image and text information, which is more challenging than the problem of only text-based conversations because it must be solved by understanding the relationship between image and text. Therefore, since there is a limit to solving only text models such as BERT or GPT2, we propose a multimodal model combining image and text. We first pretrain the multimodal model to understand the relationship between image and text, then finetune our model for each task. We achieve the 3rd best performance in subtask \#1, \#2 and a runner-up in the generation of subtask \#4. The source code is available at https://github.com/rungjoo/simmc2.0.