Learning a Text-Video Embedding from Incomplete and Heterogeneous Data
This work addresses a data scarcity problem for researchers and practitioners in video-language understanding, offering an incremental improvement over existing embedding methods.
The paper tackles the lack of large-scale annotated video-caption datasets by learning text-video embeddings from heterogeneous data sources, resulting in significant improvements on video retrieval tasks, outperforming previous methods on MPII Movie Description and MSR-VTT datasets.
Joint understanding of video and language is an active research area with many applications. Prior work in this domain typically relies on learning text-video embeddings. One difficulty with this approach, however, is the lack of large-scale annotated video-caption datasets for training. To address this issue, we aim at learning text-video embeddings from heterogeneous data sources. To this end, we propose a Mixture-of-Embedding-Experts (MEE) model with ability to handle missing input modalities during training. As a result, our framework can learn improved text-video embeddings simultaneously from image and video datasets. We also show the generalization of MEE to other input modalities such as face descriptors. We evaluate our method on the task of video retrieval and report results for the MPII Movie Description and MSR-VTT datasets. The proposed MEE model demonstrates significant improvements and outperforms previously reported methods on both text-to-video and video-to-text retrieval tasks. Code is available at: https://github.com/antoine77340/Mixture-of-Embedding-Experts