SeqDialN: Sequential Visual Dialog Networks in Joint Visual-Linguistic Representation Space
This work addresses the problem of multimodal reasoning in visual dialog for AI systems, presenting an incremental improvement over existing methods.
The paper tackles the visual dialog task by formulating it as a sequence problem using joint visual-linguistic representations, achieving state-of-the-art results with a generative model scoring 63.78% NDCG and 49.98% MRR, and a discriminative model reaching 72.41% NDCG and 55.11% MRR on the VisDial v1.0 dataset.
In this work, we formulate a visual dialog as an information flow in which each piece of information is encoded with the joint visual-linguistic representation of a single dialog round. Based on this formulation, we consider the visual dialog task as a sequence problem consisting of ordered visual-linguistic vectors. For featurization, we use a Dense Symmetric Co-Attention network as a lightweight vison-language joint representation generator to fuse multimodal features (i.e., image and text), yielding better computation and data efficiencies. For inference, we propose two Sequential Dialog Networks (SeqDialN): the first uses LSTM for information propagation (IP) and the second uses a modified Transformer for multi-step reasoning (MR). Our architecture separates the complexity of multimodal feature fusion from that of inference, which allows simpler design of the inference engine. IP based SeqDialN is our baseline with a simple 2-layer LSTM design that achieves decent performance. MR based SeqDialN, on the other hand, recurrently refines the semantic question/history representations through the self-attention stack of Transformer and produces promising results on the visual dialog task. On VisDial v1.0 test-std dataset, our best single generative SeqDialN achieves 62.54% NDCG and 48.63% MRR; our ensemble generative SeqDialN achieves 63.78% NDCG and 49.98% MRR, which set a new state-of-the-art generative visual dialog model. We fine-tune discriminative SeqDialN with dense annotations and boost the performance up to 72.41% NDCG and 55.11% MRR. In this work, we discuss the extensive experiments we have conducted to demonstrate the effectiveness of our model components. We also provide visualization for the reasoning process from the relevant conversation rounds and discuss our fine-tuning methods. Our code is available at https://github.com/xiaoxiaoheimei/SeqDialN