Learning Open-World Object Proposals Without Learning to Classify

Object proposals have become an integral pre-processing step of many vision pipelines including object detection, weakly supervised detection, object discovery, tracking, etc. Compared to the learning-free methods, learning-based proposals have become popular recently due to the growing interest in object detection. The common paradigm is to learn object proposals from data labeled with a set of object regions and their corresponding categories. However, this approach often struggles with novel objects in the open world that are absent in the training set. In this letter, we identify that the problem is that the binary classifiers in existing proposal methods tend to overfit to the training categories. Therefore, we propose a <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">classification-free</b> Object Localization Network (OLN) which estimates the objectness of each region purely by how well the location and shape of a region overlap with any ground-truth object (e.g., centerness and IoU). This strategy learns generalizable objectness and outperforms existing proposals on cross-category generalization on COCO. We further explore more challenging cross-dataset generalization onto RoboNet and EpicKitchens dataset, and long-tail detection on LVIS dataset. We demonstrate clear improvement over the state-of-the-art object detectors and object proposers. The code is publicly available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/mcahny/object_localization_network</uri> .