Learning to Compose Topic-Aware Mixture of Experts for Zero-Shot Video Captioning
arXiv:1811.02765
Abstract
Although promising results have been achieved in video captioning, existing models are limited to the fixed inventory of activities in the training corpus, and do not generalize to open vocabulary scenarios. Here we introduce a novel task, zero-shot video captioning, that aims at describing out-of-domain videos of unseen activities. Videos of different activities usually require different captioning strategies in many aspects, i.e. word selection, semantic construction, and style expression etc, which poses a great challenge to depict novel activities without paired training data. But meanwhile, similar activities share some of those aspects in common. Therefore, We propose a principled Topic-Aware Mixture of Experts (TAMoE) model for zero-shot video captioning, which learns to compose different experts based on different topic embeddings, implicitly transferring the knowledge learned from seen activities to unseen ones. Besides, we leverage external topic-related text corpus to construct the topic embedding for each activity, which embodies the most relevant semantic vectors within the topic. Empirical results not only validate the effectiveness of our method in utilizing semantic knowledge for video captioning, but also show its strong generalization ability when describing novel activities.
arXiv commentAccepted to AAAI 2019
References in corpus (14)
- ADADELTA: An Adaptive Learning Rate Method
- The Kinetics Human Action Video Dataset
- Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- Regularizing and Optimizing LSTM Language Models
- Temporal Segment Networks: Towards Good Practices for Deep Action Recognition
- Coherent Multi-Sentence Video Description with Variable Level of Detail
- Sequence to Sequence -- Video to Text
- Describing Videos by Exploiting Temporal Structure
- Decoupled Novel Object Captioner
- Deep Mixture of Experts via Shallow Embedding
- Universal Neural Machine Translation for Extremely Low Resource Languages
- S3D: Single Shot multi-Span Detector via Fully 3D Convolutional Networks
- Weakly Supervised Dense Video Captioning
- Network of Experts for Large-Scale Image Categorization