Picture for Yuexian Zou

Yuexian Zou

FiTs: Fine-grained Two-stage Training for Knowledge-aware Question Answering

Mar 15, 2023
Qichen Ye, Bowen Cao, Nuo Chen, Weiyuan Xu, Yuexian Zou

Figure 1 for FiTs: Fine-grained Two-stage Training for Knowledge-aware Question Answering

Figure 2 for FiTs: Fine-grained Two-stage Training for Knowledge-aware Question Answering

Figure 3 for FiTs: Fine-grained Two-stage Training for Knowledge-aware Question Answering

Figure 4 for FiTs: Fine-grained Two-stage Training for Knowledge-aware Question Answering

* Accepted in AAAI 2023, oral

Access Paper or Ask Questions

PoseRAC: Pose Saliency Transformer for Repetitive Action Counting

Mar 15, 2023
Ziyu Yao, Xuxin Cheng, Yuexian Zou

Figure 1 for PoseRAC: Pose Saliency Transformer for Repetitive Action Counting

Figure 2 for PoseRAC: Pose Saliency Transformer for Repetitive Action Counting

Figure 3 for PoseRAC: Pose Saliency Transformer for Repetitive Action Counting

Figure 4 for PoseRAC: Pose Saliency Transformer for Repetitive Action Counting

* 10 pages, 7 figures

Access Paper or Ask Questions

Improving Text-Audio Retrieval by Text-aware Attention Pooling and Prior Matrix Revised Loss

Mar 14, 2023
Yifei Xin, Dongchao Yang, Yuexian Zou

Figure 1 for Improving Text-Audio Retrieval by Text-aware Attention Pooling and Prior Matrix Revised Loss

Figure 2 for Improving Text-Audio Retrieval by Text-aware Attention Pooling and Prior Matrix Revised Loss

Figure 3 for Improving Text-Audio Retrieval by Text-aware Attention Pooling and Prior Matrix Revised Loss

Access Paper or Ask Questions

ZeroNLG: Aligning and Autoencoding Domains for Zero-Shot Multimodal and Multilingual Natural Language Generation

Mar 11, 2023
Bang Yang, Fenglin Liu, Yuexian Zou, Xian Wu, Yaowei Wang, David A. Clifton

Figure 1 for ZeroNLG: Aligning and Autoencoding Domains for Zero-Shot Multimodal and Multilingual Natural Language Generation

Figure 2 for ZeroNLG: Aligning and Autoencoding Domains for Zero-Shot Multimodal and Multilingual Natural Language Generation

Figure 3 for ZeroNLG: Aligning and Autoencoding Domains for Zero-Shot Multimodal and Multilingual Natural Language Generation

Figure 4 for ZeroNLG: Aligning and Autoencoding Domains for Zero-Shot Multimodal and Multilingual Natural Language Generation

* We will release the codes and models at https://github.com/yangbang18/ZeroNLG soon. Without any labeled downstream pairs for training, the ZeroNLG can deal with multiple NLG tasks, including image-to-text, video-to-text, and text-to-text, across English, Chinese, German, and French within a unified framework

Access Paper or Ask Questions

Improving Weakly Supervised Sound Event Detection with Causal Intervention

Mar 10, 2023
Yifei Xin, Dongchao Yang, Fan Cui, Yujun Wang, Yuexian Zou

Figure 1 for Improving Weakly Supervised Sound Event Detection with Causal Intervention

Figure 2 for Improving Weakly Supervised Sound Event Detection with Causal Intervention

Figure 3 for Improving Weakly Supervised Sound Event Detection with Causal Intervention

* Accepted by ICASSP2023

Access Paper or Ask Questions

FTM: A Frame-level Timeline Modeling Method for Temporal Graph Representation Learning

Feb 23, 2023
Bowen Cao, Qichen Ye, Weiyuan Xu, Yuexian Zou

Figure 1 for FTM: A Frame-level Timeline Modeling Method for Temporal Graph Representation Learning

Figure 2 for FTM: A Frame-level Timeline Modeling Method for Temporal Graph Representation Learning

Figure 3 for FTM: A Frame-level Timeline Modeling Method for Temporal Graph Representation Learning

Figure 4 for FTM: A Frame-level Timeline Modeling Method for Temporal Graph Representation Learning

* Accepted in AAAI 2023, oral

Access Paper or Ask Questions

SSVMR: Saliency-based Self-training for Video-Music Retrieval

Feb 18, 2023
Xuxin Cheng, Zhihong Zhu, Hongxiang Li, Yaowei Li, Yuexian Zou

Figure 1 for SSVMR: Saliency-based Self-training for Video-Music Retrieval

Figure 2 for SSVMR: Saliency-based Self-training for Video-Music Retrieval

Figure 3 for SSVMR: Saliency-based Self-training for Video-Music Retrieval

Figure 4 for SSVMR: Saliency-based Self-training for Video-Music Retrieval

* Accepted by ICASSP 2023

Access Paper or Ask Questions