Meta-learning for few-shot open task recognition.

Current few-shot learning research often assumes predefined task configurations and evaluates under fixed N-way K-shot settings. In realistic deployments, the target configuration is unknown at training time, and both way and shot can differ from the…

Preview only. Read the full abstract at the source