Prototype-based embedding network for scene graph generation
Abstract Current Scene Graph Generation (SGG) methods explore contextual information to
predict relationships among entity pairs. However, due to the diverse visual appearance of …
predict relationships among entity pairs. However, due to the diverse visual appearance of …
From global to local: Multi-scale out-of-distribution detection
Out-of-distribution (OOD) detection aims to detect “unknown” data whose labels have not
been seen during the in-distribution (ID) training process. Recent progress in representation …
been seen during the in-distribution (ID) training process. Recent progress in representation …
Memory-based augmentation network for video captioning
Video captioning focuses on generating natural language descriptions according to the
video content. Existing works mainly explore this multimodal learning with the paired source …
video content. Existing works mainly explore this multimodal learning with the paired source …
Learning visual question answering on controlled semantic noisy labels
Abstract Visual Question Answering (VQA) has made great progress recently due to the
increasing ability to understand and encode multi-modal inputs based on deep learning …
increasing ability to understand and encode multi-modal inputs based on deep learning …
Complementarity-aware space learning for video-text retrieval
In general, videos are powerful at recording physical patterns (eg, spatial layout) while texts
are great at describing abstract symbols (eg, emotion). When video and text are used in …
are great at describing abstract symbols (eg, emotion). When video and text are used in …
Spatial-temporal knowledge-embedded transformer for video scene graph generation
Video scene graph generation (VidSGG) aims to identify objects in visual scenes and infer
their relationships for a given video. It requires not only a comprehensive understanding of …
their relationships for a given video. It requires not only a comprehensive understanding of …
End-to-end pre-training with hierarchical matching and momentum contrast for text-video retrieval
W Shen, J Song, X Zhu, G Li… - IEEE Transactions on …, 2023 - ieeexplore.ieee.org
Lately, video-language pre-training and text-video retrieval have attracted significant
attention with the explosion of multimedia data on the Internet. However, existing …
attention with the explosion of multimedia data on the Internet. However, existing …
A differentiable semantic metric approximation in probabilistic embedding for cross-modal retrieval
Cross-modal retrieval aims to build correspondence between multiple modalities by learning
a common representation space. Typically, an image can match multiple texts semantically …
a common representation space. Typically, an image can match multiple texts semantically …
Reducing vision-answer biases for multiple-choice VQA
Multiple-choice visual question answering (VQA) is a challenging task due to the
requirement of thorough multimodal understanding and complicated inter-modality …
requirement of thorough multimodal understanding and complicated inter-modality …
Utilizing greedy nature for multimodal conditional image synthesis in transformers
Multimodal Conditional Image Synthesis (MCIS) aims to generate images according to
different modalities input and their combination, which allows users to describe their …
different modalities input and their combination, which allows users to describe their …