Self-supervised Hypergraphs for Learning Multiple World Interpretations
This work addresses the challenge of multi-task learning in computer vision for applications like autonomous systems, though it appears incremental by building on existing graph-based methods.
The paper tackles the problem of learning multiple scene representations from limited labeled data by using a multi-task hypergraph to exploit relationships between representations, achieving superior performance compared to other multi-task graph models. It also introduces Dronescapes, a large UAV video dataset with multiple representations for multi-task learning.
We present a method for learning multiple scene representations given a small labeled set, by exploiting the relationships between such representations in the form of a multi-task hypergraph. We also show how we can use the hypergraph to improve a powerful pretrained VisTransformer model without any additional labeled data. In our hypergraph, each node is an interpretation layer (e.g., depth or segmentation) of the scene. Within each hyperedge, one or several input nodes predict the layer at the output node. Thus, each node could be an input node in some hyperedges and an output node in others. In this way, multiple paths can reach the same node, to form ensembles from which we obtain robust pseudolabels, which allow self-supervised learning in the hypergraph. We test different ensemble models and different types of hyperedges and show superior performance to other multi-task graph models in the field. We also introduce Dronescapes, a large video dataset captured with UAVs in different complex real-world scenes, with multiple representations, suitable for multi-task learning.