Multimodal Transformer Networks for End-to-End Video-Grounded Dialogue Systems
By Hung Le, Doyen Sahoo, Nancy F. Chen, Steven C. H. Hoi. 2019
123 citations recorded by the source corpus.
Developing Video-Grounded Dialogue Systems (VGDS), where a dialogue is conducted based on visual and audio aspects of a given video, is significantly more challenging than traditional image or text-grounded dialogue systems because (1) feature space of videos span across multiple picture frames, making it difficult to obtain semantic information; and (2) a dialogue agent must perceive and process information from different modalities (audio, video, caption, etc.) to obtain a comprehensive understanding. Most existing work is based on RNNs and sequence-to-sequence architectures, which are not very effective for capturing complex long-term dependencies (like in videos). To overcome this, we propose Multimodal Transformer Networks (MTN) to encode videos and incorporate information from different modalities. We also propose query-aware attention through an auto-encoder to extract query-aware features from non-text modalities. We develop a training procedure to simulate token-level decoding to improve the quality of generated responses during inference. We get state of the art performance on Dialogue System Technology Challenge 7 (DSTC7). Our model also generalizes to another multimodal visual-grounded dialogue task, and obtains promising performance.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Multimodal Learning, Dialogue & Conversational AI
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Christoph Feichtenhofer
- Lorenzo Torresani
- Trevor Darrell
- Mohit Bansal
- Yanghao Li
- Kristen Grauman
- Jitendra Malik
- Cordelia Schmid
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Jitendra Malik
- Dacheng Tao
- Trevor Darrell
- Xiaogang Wang
- Ming-Hsuan Yang
- Cordelia Schmid
- Andrew Zisserman
In Multimodal Learning
- Shih-Fu Chang
- Christoph Feichtenhofer
- Lorenzo Torresani
- Haoqi Fan
- Xudong Lin
- Erik Cambria
- Bo Xiong
- Tom Young
Citations in the corpus
- Is Space-Time Attention All You Need for Video Understanding? (2021, background): “We also note that there is a wide literature based on the use of text Transformers combined with video CNNs to address various video-language tasks, such as captioning (Zhou et al., 2018), question-answering (Yang et al., 2020) and dialog (Le et al., 2019).”
- Multiscale Vision Transformers (2021, background): “An emerging thread of work aims at applying transformers to vision tasks such as object detection [5], semantic segmentation [121, 105], 3D reconstruction [77], pose estimation [113], generative modeling [17], image retrieval [30], medical image segmentation [16,103,117], point clouds [45], video in”
- Span-based Localizing Network for Natural Language Video Localization (2020, background): “As an important vision-language understanding task, NLVL involves both computer vision and natural language processing techniques (Kr-ishna et al., 2017; Hendricks et al., 2017; Gao et al., 2018; Le et al., 2019; Yu et al., 2019).”
- A Bichannel Transformer with Context Encoding for Document-Driven Conversation Generation in Social Media (2020, background): “In these works, external knowledge is leveraged to facilitate conversation understanding and generation [18, 19], which includes structured data such as knowledge graphs [20, 21], unstructured textual knowledge [22], and visual knowledge [23].”
- VX2TEXT: End-to-End Learning of Video-Based Text Generation From Multimodal Inputs (2021, methodology): “For generative tasks, e.g., captioning and video dialog, we follow previous works and use Beam Search [41, 25] (with beam width set to 5) or Greedy Decoding [28] to generate coherent sentences.”
- Recent advances in deep learning based dialogue systems: a systematic survey (2021, methodology): “There are some recent works trying to build multimodal dialogue systems (Le et al., 2019; Chauhan et al., 2019; Saha et al., 2020; Singla et al., 2020; Young et al., 2020), but these systems are still far from mature.”