What Object Should I Use? - Task Driven Object Detection
This addresses a gap in object detection benchmarks for task-driven object selection, which is crucial for robotics and autonomous systems, representing a novel domain-specific advancement.
The paper tackles the problem of selecting the most suitable objects for specific tasks, which is challenging for autonomous systems, by introducing the COCO-Tasks dataset with 40,000 images annotated for 14 tasks and proposing a Gated Graph Neural Network approach that outperforms other methods on this dataset.
When humans have to solve everyday tasks, they simply pick the objects that are most suitable. While the question which object should one use for a specific task sounds trivial for humans, it is very difficult to answer for robots or other autonomous systems. This issue, however, is not addressed by current benchmarks for object detection that focus on detecting object categories. We therefore introduce the COCO-Tasks dataset which comprises about 40,000 images where the most suitable objects for 14 tasks have been annotated. We furthermore propose an approach that detects the most suitable objects for a given task. The approach builds on a Gated Graph Neural Network to exploit the appearance of each object as well as the global context of all present objects in the scene. In our experiments, we show that the proposed approach outperforms other approaches that are evaluated on the dataset like classification or ranking approaches.