Few-Shot Object Detection
39 papers with code • 6 benchmarks • 5 datasets
Target: To detect objects of novel categories with just a few training samples.
A clear explanation of the few-shot object detection task and its differences with few-shot classification can be found in "A Survey of Self-Supervised and Few-Shot Object Detection": https://gabrielhuang.github.io/fsod-survey/
LibrariesUse these libraries to find Few-Shot Object Detection models and implementations
We demonstrate empirical results on MS Coco highlighting challenges of the one-shot setting: while transferring knowledge about instance segmentation to novel object categories works very well, targeting the detection network towards the reference category appears to be more difficult.
To train our network, we contribute a new dataset that contains 1000 categories of various objects with high-quality annotations.
Few-shot object detection (FSOD) helps detectors adapt to unseen classes with few training instances, and is useful when manual annotation is time-consuming or data acquisition is limited.
Few-shot object detection has made substantial progressby representing novel class objects using the feature representation learned upon a set of base class objects.
Few-shot object detection has been extensively investigated by incorporating meta-learning into region-based detection frameworks.
To improve the fine-grained few-shot proposal classification, we propose a novel attentive feature alignment method to address the spatial misalignment between the noisy proposals and few-shot classes, thus improving the performance of few-shot object detection.
Second, we introduce a novel regularized transfer learning framework for low-shot detection, where the transfer knowledge (TK) and background depression (BD) regularizations are proposed to leverage object knowledge respectively from source and target domains, in order to further enhance fine-tuning with a few target images.