Multi-Modal Fusion by Meta-Initialization
When experience is scarce, models may have insufficient information to adapt to a new task. In this case, auxiliary information - such as a textual description of the task - can enable improved task inference and adaptation. In this work, we propose an extension to the Model-Agnostic Meta-Learning algorithm (MAML), which allows the model to adapt using auxiliary information as well as task experience. Our method, Fusion by Meta-Initialization (FuMI), conditions the model initialization on auxiliary information using a hypernetwork, rather than learning a single, task-agnostic initialization. Furthermore, motivated by the shortcomings of existing multi-modal few-shot learning benchmarks, we constructed iNat-Anim - a large-scale image classification dataset with succinct and visually pertinent textual class descriptions. On iNat-Anim, FuMI significantly outperforms uni-modal baselines such as MAML in the few-shot regime. The code for this project and a dataset exploration tool for iNat-Anim are publicly available at https://github.com/s-a-malik/multi-few .
Code (1)
Tasks
Few-Shot Learningimage-classificationImage ClassificationMeta-LearningMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Dynamic Multimodal Fusion via Meta-Learning Towards Micro-Video Recommendation
Multimodal information (e.g., visual, acoustic, and textual) has been widely used to enhance representation learning for micro-video recommendation. For integrating multimodal information into a joint representation of m…
Meta-LearningMultimodal RecommendationRepresentation LearningToward Multimodal Model-Agnostic Meta-Learning
Gradient-based meta-learners such as MAML are able to learn a meta-prior from similar tasks to adapt to novel tasks from the same distribution with few gradient updates. One important limitation of such frameworks is tha…
DiversityFew-Shot Image Classificationimage-classificationImage Classification+3Model-Agnostic Meta-Learning for Multimodal Task Distributions
Gradient-based meta-learners such as MAML (Finn et al., 2017) are able to learn a meta-prior from similar tasks to adapt to novel tasks from the same distribution with few gradient updates. One important limitation of su…
DiversityFew-Shot Image Classificationimage-classificationImage Classification+2MSNN-LINet: Cross-Modal Learning via Continuous Linear Integration
We present LINet (Linear Integration Network), a Multi-Stream Neural Network (MSNN) for RGB-D scene classification. Current multi-modal architectures treat feature fusion as a discrete, ad-hoc event: early fusion entangl…
Scene ClassificationMultimodal Model-Agnostic Meta-Learning via Task-Aware Modulation
Model-agnostic meta-learners aim to acquire meta-learned parameters from similar tasks to adapt to novel tasks from the same distribution with few gradient updates. With the flexibility in the choice of models, those fra…
Few-Shot Image ClassificationFew-Shot LearningGeneral Classificationimage-classification+5