Image Co-Saliency Detection and Instance Co-Segmentation Using Attention Graph Clustering Based Graph Convolutional Network

Li, Tengpeng; Zhang, Kaihua; Shen, Shiwen; Liu, Bo; Liu, Qingshan; Zhu, Li

doi:10.1109/tmm.2021.3054526

Cited by 31 publications

(11 citation statements)

References 91 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, modified GNN architectures have been developed for 3D instance segmentation on point clouds. There have been two main approaches: predicting bounding shapes to separate individual object instances after re-embedding the graph through edge and/or node convolutions [142,143] and (possibly attention based) graph pooling/clustering/condensation networks followed by localized node classification [144][145][146].…”

Section: Instance Segmentation Approachesmentioning

confidence: 99%

Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges

Thais¹,

Calafiura²,

Chachamis³

et al. 2022

Preprint

View full text Add to dashboard Cite

Many physical systems can be best understood as sets of discrete data with associated relationships. Where previously these sets of data have been formulated as series or image data to match the available machine learning architectures, with the advent of graph neural networks (GNNs), these systems can be learned natively as graphs. This allows a wide variety of high-and low-level physical features to be attached to measurements and, by the same token, a wide variety of HEP tasks to be accomplished by the same GNN architectures. GNNs have found powerful use-cases in reconstruction, tagging, generation and end-to-end analysis. With the wide-spread adoption of GNNs in industry, the HEP community is well-placed to benefit from rapid improvements in GNN latency and memory usage. However, industry use-cases are not perfectly aligned with HEP and much work needs to be done to best match unique GNN capabilities to unique HEP obstacles. We present here a range of these capabilities,

show abstract

Section: Instance Segmentation Approachesmentioning

confidence: 99%

Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges

Thais¹,

Calafiura²,

Chachamis³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Recent deep-based methods attempt to detect co-salient objects with an end-to-end manner. Li et al [7] introduce an attention graph clustering model to graph neural networks to explore the group semantics. Zhang et al [6] extract group semantics by the pre-training classification features and calibrate the individuals with feedback gradient information.…”

Section: Related Workmentioning

confidence: 99%

“…This means that the learned group semantics are scattered and vulnerable, and it is easy to produce less robust performance in the inference. On the other hand, previous methods [6,7,8,9] ignore the distinction between categories or completely negate the correlation between categories [5], which fails to suppress noise objects. The former pays more attentions to the positive relations of intra-group repetitive objects and lacks the learning of the negative relations among inter-group objects, which leads to the lack of outlier object detection.…”

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

A Unified Two-Stage Group Semantics Propagation and Contrastive Learning Network for Co-Saliency Detection

Tan¹,

Chen²,

Wen³

et al. 2022

Preprint

View full text Add to dashboard Cite

Co-saliency detection (CoSOD) aims at discovering the repetitive salient objects from multiple images. Two primary challenges are group semantics extraction and noise object suppression. In this paper, we present a unified Two-stage grOup semantics PropagatIon and Contrastive learning NETwork (TopicNet) for CoSOD. TopicNet can be decomposed into two substructures, including a two-stage group semantics propagation module (TGSP) to address the first challenge and a contrastive learning module (CLM) to address the second challenge. Concretely, for TGSP, we design an imageto-group propagation module (IGP) to capture the consensus representation of intra-group similar features and a group-topixel propagation module (GPP) to build the relevancy of consensus representation. For CLM, with the design of positive samples, the semantic consistency is enhanced. With the design of negative samples, the noise objects are suppressed. Experimental results on three prevailing benchmarks reveal that TopicNet outperforms other competitors in terms of various evaluation metrics.

show abstract

“…Additionally, all images in GDXray are gray-scale, and less than 1% pictures in SIXray are annotated with prohibited items, both of which are not of very high quality. In addition, traditional CNN-based models [8]- [13] trained through common image recognition datasets fail to achieve satisfactory performance in this scenario. This urgently requires researchers to make breakthroughs in both datasets and models.…”

Section: Introductionmentioning

confidence: 99%

Over-sampling De-occlusion Attention Network for Prohibited Items Detection in Noisy X-ray Images

Tao,

Wei,

et al. 2021

Preprint

View full text Add to dashboard Cite

Security inspection is X-ray scanning for personal belongings in suitcases, which is significantly important for the public security but highly time-consuming for human inspectors. Fortunately, deep learning has greatly promoted the development of computer vision, offering a possible way of automatic security inspection. However, items within a luggage are randomly overlapped resulting in noisy X-ray images with heavy occlusions. Thus, traditional CNN-based models trained through common image recognition datasets fail to achieve satisfactory performance in this scenario.To address these problems, we contribute the first high-quality prohibited X-ray object detection dataset named OPIXray, which contains 8885 X-ray images from 5 categories of the widely-occurred prohibited item "cutters". The images are gathered from an airport and these prohibited items are annotated manually by professional inspectors, which can be used as a benchmark for model training and further facilitate future research. To better improve occluded X-ray object detection, we further propose an over-sampling deocclusion attention network (DOAM-O), which consists of a novel de-occlusion attention module and a new over-sampling training strategy. Specifically, our de-occlusion module, namely DOAM, simultaneously leverages the different appearance information of the prohibited items; the over-sampling training strategy forces the model to put more emphasis on these hard samples consisting these items of high occlusion levels, which is more suitable for this scenario. We comprehensively evaluated DOAM-O on the OPIXray dataset, which proves that our model can stably improve the performance of the famous detection models such as SSD, YOLOv3, and FCOS, and outperform many extensivelyused attention mechanisms. 1

show abstract

Image Co-Saliency Detection and Instance Co-Segmentation Using Attention Graph Clustering Based Graph Convolutional Network

Cited by 31 publications

References 91 publications

Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges

Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges

A Unified Two-Stage Group Semantics Propagation and Contrastive Learning Network for Co-Saliency Detection

Over-sampling De-occlusion Attention Network for Prohibited Items Detection in Noisy X-ray Images

Contact Info

Product

Resources

About