Towards Accurate Localization by Instance Search
This work addresses the challenge of localizing objects of unknown categories for computer vision applications, offering a novel approach that is not incremental but builds on weakly supervised methods.
The paper tackles the problem of visual object localization without object-level annotations by proposing a self-paced learning framework that mines target instances from instance search results, achieving superior performance over state-of-the-art methods on localization and few-shot object detection tasks.
Visual object localization is the key step in a series of object detection tasks. In the literature, high localization accuracy is achieved with the mainstream strongly supervised frameworks. However, such methods require object-level annotations and are unable to detect objects of unknown categories. Weakly supervised methods face similar difficulties. In this paper, a self-paced learning framework is proposed to achieve accurate object localization on the rank list returned by instance search. The proposed framework mines the target instance gradually from the queries and their corresponding top-ranked search results. Since a common instance is shared between the query and the images in the rank list, the target visual instance can be accurately localized even without knowing what the object category is. In addition to performing localization on instance search, the issue of few-shot object detection is also addressed under the same framework. Superior performance over state-of-the-art methods is observed on both tasks.