Locating Objects Without Bounding Boxes
This addresses the labeling bottleneck in object detection for applications like tracking and monitoring, though it is an incremental improvement over existing localization techniques.
The paper tackles the problem of object localization without requiring bounding box annotations, which are time-consuming to label, by proposing a loss function based on the average Hausdorff distance for use in fully convolutional networks. The method outperforms state-of-the-art generic object detectors and fine-tuned methods on datasets for locating heads, pupil centers, and plant centers.
Recent advances in convolutional neural networks (CNN) have achieved remarkable results in locating objects in images. In these networks, the training procedure usually requires providing bounding boxes or the maximum number of expected objects. In this paper, we address the task of estimating object locations without annotated bounding boxes which are typically hand-drawn and time consuming to label. We propose a loss function that can be used in any fully convolutional network (FCN) to estimate object locations. This loss function is a modification of the average Hausdorff distance between two unordered sets of points. The proposed method has no notion of bounding boxes, region proposals, or sliding windows. We evaluate our method with three datasets designed to locate people's heads, pupil centers and plant centers. We outperform state-of-the-art generic object detectors and methods fine-tuned for pupil tracking.