OS2D: One-Stage One-Shot Object Detection by Matching Anchor Features

Osokin, Anton; Sumin, Denis; Lomakin, Vasily

doi:10.1007/978-3-030-58555-6_38

Cited by 38 publications

(29 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A common framework can be observed in most of the works in this section. In fact, in [13,19,20,27,35] the query image and the support image(s) traverse two parallel branches. The support branch is supposed to extract meaningful features from the samples of each class to instill them into the result of the query branch by aggregation.…”

Section: Discussionmentioning

confidence: 99%

“…Moreover, as for many recent works, a crucial role is played by attention, which is used in [13,19,20]. Regarding the architecture, Faster R-CNN is still the preferred choice, with only [27,35] employing a one-stage detector. Finally, a meaningful and discriminative embedding space is the key of the success of metric learning approaches: this is achieved in [20,27,47] by employing a margin-based loss and in [13] by introducing a contrastive training scheme.…”

Section: Discussionmentioning

confidence: 99%

“…The other approaches, instead, must be separately treated. We can notice that the tasks employed in [8], [54], [47], [20], [35], and [13] are diferent from the standard ones. In fact, in Table 3 NS1, NS2, and N3 are replaced with custom tasks (T1, T2, T3, C1, C2, C3, and C4).…”

Section: Summary Table and Discussionmentioning

confidence: 99%

“…). [35] The authors propose a novel model called OS2D, based on a TransformNet especially focused on One-Shot OD, in which the dense grid of anchor locations are taken together from a Faster R-CNN and a SSD. The proposal consists of a 4-step pipeline: features extraction, correlation matching, spatial alignment, and computing outputs.…”

Section: Os2d: One-stage One-shot Object Detection By Matching Anchor...mentioning

confidence: 99%

See 3 more Smart Citations

Few-Shot Object Detection: A Survey

et al. 2022

View full text Add to dashboard Cite

Deep Learning approaches have recently raised the bar in many fields, from Natural Language Processing to Computer Vision, by leveraging large amounts of data. However, they could fail when the retrieved information is not enough to fit the vast number of parameters, frequently resulting in overfitting and, therefore, in poor generalizability. Few-Shot Learning aims at designing models which can effectively operate in a scarce data regime, yielding learning strategies that only need few supervised examples to be trained. These procedures are of both practical and theoretical importance, as they are crucial for many real-life scenarios in which data is either costly or even impossible to retrieve. Moreover, they bridge the distance between current data-hungry models and human-like generalization capability. Computer Vision offers various tasks which can be few-shot inherent, such as person re-identification. This survey, which to the best of our knowledge is the first tackling this problem, is focused on Few-Shot Object Detection, which has received far less attention compared to Few-Shot Classification due to the intrinsic challenge level. In this regard, this review presents an extensive description of the approaches that have been tested in the current literature, discussing their pros and cons, and classifying them according to a rigorous taxonomy.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Summary Table and Discussionmentioning

confidence: 99%

Section: Os2d: One-stage One-shot Object Detection By Matching Anchor...mentioning

confidence: 99%

See 2 more Smart Citations

Few-Shot Object Detection: A Survey

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Meta-learner [16,3,29,43] is introduced to acquire class-level meta knowledge via feature re-weighting and helps the model to generalize to novel classes. Metalearning based methods [39,18,41,3,44,7,14,28] has been demonstrated to be successful for FSOD. Moreover, meta-learning based methods can be efficient for incrementally adding new few-shot classes during network inference.…”

Section: Related Workmentioning

confidence: 99%

Query Adaptive Few-Shot Object Detection with Heterogeneous Graph Convolutional Networks

Han¹,

He²,

Huang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Few-shot object detection (FSOD) aims to detect neverseen objects using few examples. This field sees recent improvement owing to the meta-learning techniques by learning how to match between the query image and few-shot class examples, such that the learned model can generalize to few-shot novel classes. However, currently, most of the meta-learning-based methods perform pairwise matching between query image regions (usually proposals) and novel classes separately, therefore failing to take into account multiple relationships among them. In this paper, we propose a novel FSOD model using heterogeneous graph convolutional networks. Through efficient message passing among all the proposal and class nodes with three different types of edges, we could obtain context-aware proposal features and query-adaptive, multiclass-enhanced prototype representations for each class, which could help promote the pairwise matching and improve final FSOD accuracy. Extensive experimental results show that our proposed model, denoted as QA-FewDet, outperforms the current state-of-theart approaches on the PASCAL VOC and MSCOCO FSOD benchmarks under different shots and evaluation metrics.

show abstract