Detailed feature extraction network-based fine-grained face segmentation

Umirzakova, Sabina; Whangbo, Taeg Keun

doi:10.1016/j.knosys.2022.109036

Cited by 17 publications

(3 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…8. Aspect ratio is often used to obtain human characteristics [51]. Human faces also have a feature that is basically longer in the vertical direction than in the horizontal direction, as illustrated in Fig.…”

Section: Human Detection Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

Autonomous Movement Control of Coaxial Mobile Robot based on Aspect Ratio of Human Face for Public Relation Activity Using Stereo Thermal Camera

Rusydi

Novira

Nakagome

et al. 2022

Journal of Robotics and Control

View full text Add to dashboard Cite

In recent years, robots that recognize people around them and provide guidance, information, and monitoring have been attracting attention. The mainstream of conventional human recognition technology is the method using a camera or laser range finder. However, it is difficult to recognize with a camera due to fluctuations in lighting 1), and it is often affected by the recognition environment such as misrecognition 2) with a person's leg and a chair's leg with a laser range finder. Therefore, we propose a human recognition method using a thermal camera that can visualize human heat. This study aims to realize human-following autonomous movement based on human recognition. In addition, the distance from the robot to the person is measured with a stereo thermal camera that uses two thermal cameras. A coaxial two-wheeled robot that is compact and capable of super-credit turning is used as a mobile robot. Finally, we conduct an autonomous movement experiment of a coaxial mobile robot based on human recognition by combining these. We performed human-following experiments on a coaxial two-wheeled robot based on human recognition using a stereo thermal camera and confirmed that it moves appropriately to the location where the recognized person is in multiple use cases (scenarios). However, the accuracy of distance measurement by stereo vision is inferior to that of laser measurement. It is necessary to improve it in the case of movement that requires more accuracy.

show abstract

Section: Human Detection Algorithmmentioning

confidence: 99%

“…As verification experiments, we measure circularity [47] and aspect ratio [51], set parameters, and measure distance with a stereo camera. A thermal image of a human face was acquired using a thermal camera, and the circularity was measured.…”

Section: A Face Detection Using Circularity and Aspect Rationmentioning

confidence: 99%

Autonomous Movement Control of Coaxial Mobile Robot based on Aspect Ratio of Human Face for Public Relation Activity Using Stereo Thermal Camera

Rusydi

Novira

Nakagome

et al. 2022

Journal of Robotics and Control

View full text Add to dashboard Cite

show abstract

“…Deep learning-based image processing techniques show improved performance in a variety of tasks, including detection [3] and segmentation [4], which can be used to develop Remote Sens. 2022, 14, 6302 2 of 20 wildfire-and smoke-processing techniques.…”

Section: Introductionmentioning

confidence: 99%

Deep Encoder–Decoder Network-Based Wildfire Segmentation Using Drone Images in Real-Time

2022

View full text Add to dashboard Cite

Wildfire is a hazardous natural phenomenon that leads to significant human fatalities, catastrophic environmental damages, and economic losses. Over the past few years, the intensity and frequency of fires have increased worldwide. Studies have been conducted to develop distinctive solutions to minimize forest fires. Systems for distant fire detection and monitoring have been established, showing improvements in data collection and fire characterization. However, wildfires cover vast areas, making other proposed ground systems unsuitable for optimal coverage. Unmanned aerial vehicles (UAVs) have become the subject of active research in recent years. Deep learning-based image-processing methods demonstrate improved performance in various tasks, including detection and segmentation, which can be utilized to develop modern forest firefighting techniques. In this study, we established a novel two-pathway encoder–decoder-based model to detect and accurately segment wildfires and smoke from the images captured using UAVs in real-time. Our proposed nested decoder uses pre-activated residual blocks and an attention-gating mechanism, thereby improving segmentation accuracy. Moreover, to facilitate robust and generalized training, we prepared a new dataset comprising actual incidences of forest fires and smoke, varying from small to large areas. In terms of practicality, the experimental results reveal that our method significantly outperforms existing detection and segmentation methods, despite being lightweight. In addition, the proposed model is reliable and robust for detecting and segmenting drone camera images from different viewpoints in the presence of wildfire and smoke.

show abstract