Which One Are You Referring To? Multimodal Object Identification in Situated Dialogue
This addresses the challenge of interpreting multimodal inputs in dialogue systems for domains requiring situational context, representing an incremental advancement.
The paper tackles the problem of multimodal object identification in situated dialogue by exploring three methods, with the best method, scene-dialogue alignment, improving performance by approximately 20% F1-score compared to baselines on the SIMMC 2.1 dataset.
The demand for multimodal dialogue systems has been rising in various domains, emphasizing the importance of interpreting multimodal inputs from conversational and situational contexts. We explore three methods to tackle this problem and evaluate them on the largest situated dialogue dataset, SIMMC 2.1. Our best method, scene-dialogue alignment, improves the performance by ~20% F1-score compared to the SIMMC 2.1 baselines. We provide analysis and discussion regarding the limitation of our methods and the potential directions for future works. Our code is publicly available at https://github.com/holylovenia/multimodal-object-identification.