Few-Shot Learning with Metric-Agnostic Conditional Embeddings
This addresses the challenge of few-shot learning for fine-grained classification tasks, offering a flexible approach that improves performance in this domain.
The paper tackles the problem of learning class representations from few examples in few-shot learning by introducing a novel architecture that conditions class representations on a target image and trains a network to perform comparisons rather than using a static metric. It achieves state-of-the-art performance on the Caltech-UCSD birds fine-grained classification task.
Learning high quality class representations from few examples is a key problem in metric-learning approaches to few-shot learning. To accomplish this, we introduce a novel architecture where class representations are conditioned for each few-shot trial based on a target image. We also deviate from traditional metric-learning approaches by training a network to perform comparisons between classes rather than relying on a static metric comparison. This allows the network to decide what aspects of each class are important for the comparison at hand. We find that this flexible architecture works well in practice, achieving state-of-the-art performance on the Caltech-UCSD birds fine-grained classification task.