SIMMC 2.0: A Task-oriented Dialog Dataset for Immersive Multimodal Conversations

TL;DR

SIMMC 2.0 introduces a large, immersive multimodal dialog dataset in the shopping domain, enabling advanced research in context-aware task-oriented systems within realistic environments.

Contribution

The paper presents SIMMC 2.0, a novel multimodal dialog dataset with a two-phase collection process and detailed benchmarks for immersive, context-aware conversations.

Findings

Baseline models achieve promising results.

The dataset reveals new challenges for multimodal dialog understanding.

Provides a foundation for future immersive dialog system research.

Abstract

Next generation task-oriented dialog systems need to understand conversational contexts with their perceived surroundings, to effectively help users in the real-world multimodal environment. Existing task-oriented dialog datasets aimed towards virtual assistance fall short and do not situate the dialog in the user's multimodal context. To overcome, we present a new dataset for Situated and Interactive Multimodal Conversations, SIMMC 2.0, which includes 11K task-oriented user<->assistant dialogs (117K utterances) in the shopping domain, grounded in immersive and photo-realistic scenes. The dialogs are collected using a two-phase pipeline: (1) A novel multimodal dialog simulator generates simulated dialog flows, with an emphasis on diversity and richness of interactions, (2) Manual paraphrasing of the generated utterances to collect diverse referring expressions. We provide an in-depth analysis of the collected dataset, and describe in detail the four main benchmark tasks we propose. Our baseline model, powered by the state-of-the-art language model, shows promising results, and highlights new challenges and directions for the community to study.