Existing meta-learning based few-shot learning (FSL) methods typically adopt an episodic training strategy whereby each episode contains a meta-task. Across episodes, these tasks are sampled randomly and their relationships are ignored. In this paper, we argue that the inter-meta-task relationships should be exploited to learn models that are more generalizable to unseen classes with few-shots. Specifically, we consider the relationships between two types of meta-tasks and propose different strategies to exploit them. (1) Two meta-tasks with disjoint sets of classes: these are interesting because their relationship is reminiscent of that between the source seen classes and target unseen classes, featured with domain gap caused by class differences. A novel meta-training strategy named meta-domain adaptation (MDA) is proposed to make the meta-learned model more robust to the domain gap. (2) Two meta-tasks with identical sets of classes: these are interesting because they can be used to learn models that are robust against poorly sampled few-shots. To that end, a novel meta-knowledge distillation (MKD) strategy is formulated. Extensive experiments demonstrate that both MDA and MKD significantly boost the performance of a variety of existing FSL methods and thus achieve new state-of-the-art on three benchmarks.

It's called the Mine Kafon Drone (MKD) and its creators have just launched a Kickstarter campaign for its production. The MKD is a hexcopter with three interchangeable arms: a high resolution camera, a metal detector and a robotic arm. The drone first flies over the field and uses its camera to both create a 3D aerial map and mark potentially dangerous areas with GPS waypoints. It uses the 3D map it made in the previous step to keep the detector just 4 cm from the ground as it flies by. Any mines that it finds are geotagged for removal.