PSNet: Perspective-sensitive convolutional network for object detection

Zhang, Xin; Liu, Yicheng; Huo, Chunlei; Xu, Nuo; Wang, Lingfeng; Chen, Pan

doi:10.1016/j.neucom.2021.10.068

Cited by 14 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…But still, challenges affect the performance of the object detector to meet with real-time performance like human beings. Tremendous research works have been done in object detection to handle the challenges in different application areas as shown in the Table 4 [ 102 , 107 , 123 , 164 , 168 , 188 , 194 , 200 , 208 , 213 , 215 ]. For example, Li.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Tools, techniques, datasets and application areas for object detection in an image: a review

Kaur

Singh

2022

Multimed Tools Appl

View full text Add to dashboard Cite

Object detection is one of the most fundamental and challenging tasks to locate objects in images and videos. Over the past, it has gained much attention to do more research on computer vision tasks such as object classification, counting of objects, and object monitoring. This study provides a detailed literature review focusing on object detection and discusses the object detection techniques. A systematic review has been followed to summarize the current research work’s findings and discuss seven research questions related to object detection. Our contribution to the current research work is (i) analysis of traditional, two-stage, one-stage object detection techniques, (ii) Dataset preparation and available standard dataset, (iii) Annotation tools, and (iv) performance evaluation metrics. In addition, a comparative analysis has been performed and analyzed that the proposed techniques are different in their architecture, optimization function, and training strategies. With the remarkable success of deep neural networks in object detection, the performance of the detectors has improved. Various research challenges and future directions for object detection also has been discussed in this research paper.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Object detection from different view angles also is a challenging issue. Zhang et al have presented an approach PSNet (Perspective-sensitive network) in which perspective-specific structural features are used instead of uniformed features [ 213 ]…”

Section: Applications Of Object Detectionmentioning

confidence: 99%

Tools, techniques, datasets and application areas for object detection in an image: a review

Kaur

Singh

2022

Multimed Tools Appl

View full text Add to dashboard Cite

show abstract

“…At present, the interpretability of deep learning is divided into several branches, among which the visualization method is one of the important research directions. Zhang et al [38] proposed sensitivity analysis to quantify the sensitivity of the model to input variables and visualize regions with high sensitivity, indicating that this region mainly affected model decisionmaking. Other visualization methods sample the image blocks with the largest convolution kernel activation value [39], and then visualize these activated image blocks to analyze how the networks obtain information.…”

Section: A Deep Learning Interpretability Approachesmentioning

confidence: 99%

G2Grad-CAMRL: An Object Detection and Interpretation Model Based on Gradient-Weighted Class Activation Mapping and Reinforcement Learning in Remote Sensing Images

Yin

Wang

Shafiq

et al. 2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Remote sensing images contain important information such as airports, ports, and ships. By extracting remote sensing image features and learning the mapping relationship between image features and text semantic features, the interpretation and description of remote sensing image content can be realized, which has a wide range of application value in military and civil fields such as national defense security, land monitoring, urban planning, disaster mitigation and so on. Aiming at the complex background of remote sensing images and the lack of interpretability of existing target detection models, and the problems in feature extraction between different network structures, different layers, and the accuracy of target classification, we propose an object detection and interpretation model based on Gradient weighted class activation mapping and reinforcement learning. Firstly, ResNet is used as the main backbone network to extract the features of remote sensing images and generate feature graphs. Then, we add the global average pooling layer to obtain the corresponding feature weight vector of the feature graph. The weighted vectors are superimposed to output class activation maps. The reinforcement learning method is used to optimize the generated region generation network. At the same time, we improve the reward function of reinforcement learning to improve the effect of the region generation network. Finally, network dissecting analysis is used to obtain the interpretable semantic concept in the model. Through experiments, the average accuracy is more than 85%. Experimental results in the public remote sensing image description dataset show that the proposed method has high detection accuracy and good description performance for remote sensing images in complex environments.

show abstract

“…A new type of convolutional network is presented in [ 100 ]. PSnet is a perspective-sensitive network to detect objects from different perspectives (i.e., angles of view).…”

Section: Related Workmentioning

confidence: 99%

Deep Learning Methods for Space Situational Awareness in Mega-Constellations Satellite-Based Internet of Things Networks

Massimi

Ferrara²,

Benedetto

2022

Sensors

View full text Add to dashboard Cite

Artificial Intelligence of things (AIoT) is the combination of Artificial Intelligence (AI) technologies and the Internet of Things (IoT) infrastructure. AI deals with the devices’ learning process to acquire knowledge from data and experience, while IoT concerns devices interacting with each other using the Internet. AIoT has been proven to be a very effective paradigm for several existing applications as well as for new areas, especially in the field of satellite communication systems with mega-constellations. When AIoT meets space communications efficiently, we have interesting uses of AI for Satellite IoT (SIoT). In fact, the number of space debris is continuously increasing as well as the risk of space collisions, and this poses a significant threat to the sustainability and safety of space operations that must be carefully and efficiently addressed to avoid critical damage to the SIoT networks. This paper aims to provide a systematic survey of the state of the art, challenges, and perspectives on the use of deep learning methods for space situational awareness (SSA) object detection and classification. The contributions of this paper can be summarized as follows: (i) we outline using AI algorithms, and in particular, deep learning (DL) methods, the possibility of identifying the nature/type of spatial objects by processing signals from radars; (ii) we present a comprehensive taxonomy of DL-based methods applied to SSA object detection and classification, as well as their characteristics, and implementation issues.

show abstract

PSNet: Perspective-sensitive convolutional network for object detection

Cited by 14 publications

References 20 publications

Tools, techniques, datasets and application areas for object detection in an image: a review

Tools, techniques, datasets and application areas for object detection in an image: a review

G2Grad-CAMRL: An Object Detection and Interpretation Model Based on Gradient-Weighted Class Activation Mapping and Reinforcement Learning in Remote Sensing Images

Deep Learning Methods for Space Situational Awareness in Mega-Constellations Satellite-Based Internet of Things Networks

Contact Info

Product

Resources

About