IoT-Based Strawberry Disease Detection With Wall-Mounted Monitoring Cameras

Lin, Yi-Bing; Liu, Chun-You; Chen, Wen-Liang; Chang, Chia-Hui; Ng, Fung-Ling; Yang, Krista; Hsung, Jerry

doi:10.1109/jiot.2023.3288603

Cited by 5 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The AI models are packaged as IoT devices and can be easily integrated into ViolinTalk. Further details can be found in [19].…”

Section: Table 2 the Violintalk Robot Patentsmentioning

confidence: 99%

ViolinTalk: Violin Robots as Internet of Things Devices

Huang,

Lin

2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

In recent years, the application of intelligent robots in the field of music has gradually emerged. Among string instruments, the violin stands as an unparalleled musical marvel, representing the soprano voice. Countless composers have penned masterpieces for this cherished instrument. This paper proposes ViolinTalk, a system for controlling XY-type violin robots as Internet of Things (IoT) devices. ViolinTalk enhances the performance of robots through calibration and testing. ViolinTalk has improved the software of violin robots with an IoT-based modular design, resulting in a better development environment. In this environment, we have made key contributions, including: (1) The development of two control modes for the bowing servo motors, comprising displacement-based control and time-based control; (2) Utilizing bowing speed control to introduce variations in volume and tone, enriching the performance effects; (3) Analyzing errors in playing speed and establishing correction parameters for future use; (4) Analyzing the robot's mechanism to enhance its stability and durability. ViolinTalk also leverages IoT technology to facilitate seamless connections between different music instruments for ensemble performances. Multiple robots have been successfully integrated into a string orchestra, featuring two violins, one viola, and one cello, capable of performing standard chamber music. This development has resulted in the acquisition of a total of 8 patents from Taiwan.INDEX TERMS Arts and humanities, Bow speed control, Internet of Things (IoT), string control, violin robot

show abstract

“…The AI models are packaged as IoT devices and can be easily integrated into ViolinTalk. Further details can be found in [19].…”

Section: Table 2 the Violintalk Robot Patentsmentioning

confidence: 99%

ViolinTalk: Violin Robots as Internet of Things Devices

Huang,

Lin

2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…Deep learning models for moving object detection based on CNNs [8][9][10][11] can utilize camera images from specific areas as training data to learn the features of moving objects. This enables accurate identification of moving objects within the frame.…”

Section: Related Workmentioning

confidence: 99%

FenceTalk: Exploring False Negatives in Moving Object Detection

Lin,

Liu,

Lin

et al. 2023

Algorithms

Self Cite

View full text Add to dashboard Cite

Deep learning models are often trained with a large amount of labeled data to improve the accuracy for moving object detection in new fields. However, the model may not be robust enough due to insufficient training data in the new field, resulting in some moving objects not being successfully detected. Training with data that is not successfully detected by the pre-trained deep learning model can effectively improve the accuracy for the new field, but it is costly to retrieve the image data containing the moving objects from millions of images per day to train the model. Therefore, we propose FenceTalk, a moving object detection system, which compares the difference between the current frame and the background image based on the structural similarity index measure (SSIM). FenceTalk automatically selects suspicious images with moving objects that are not successfully detected by the Yolo model, so that the training data can be selected at a lower labor cost. FenceTalk can effectively define and update the background image in the field, reducing the misjudgment caused by changes in light and shadow, and selecting images containing moving objects with an optimal threshold. Our study has demonstrated its performance and generality using real data from different fields. For example, compared with the pre-trained Yolo model using the MS COCO dataset, the overall recall of FenceTalk increased from 72.36% to 98.39% for the model trained with the data picked out by SSIM. The recall of FenceTalk, combined with Yolo and SSIM, can reach more than 99%.

show abstract

“…To derive a quick and primary estimation for 𝑡 * , we conduct the mean value analysis [28], where both 𝑇 𝐴 − 𝑇 𝐴 0 and 𝑇 𝐵 − 𝑇 𝐵 0 have the means 1/𝜆 𝐴 and 1/𝜆 𝐵 , and the Exponential density functions 𝑓 𝐴 (𝑡) = 𝜆 𝐴 𝑒 −𝜆 𝐴 𝑡 and 𝑓 𝐵 (𝑡) = 𝜆 𝐵 𝑒 −𝜆 𝐵 𝑡 (6) We consider a more general scenario where Puppeteer A is replaced by 𝐼 ≥ 1 puppeteers, and Puppeteer B needs to wait for the other I puppeteers' actions before taking the next action. Suppose that the i-th puppeteer issues the action at time 𝑇 𝑖 0 and the i-th puppet/avatar performs at 𝑇 𝑖 .…”

Section: Viic Delay Time Performancementioning

confidence: 99%

An IoT-Based Microservice Platform for Virtual Puppetry Performance

Luo,

Lin,

Liao

et al. 2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

This study proposes a web-based real-time remote-controlled virtual puppetry (avatar) performance platform called AvatarTalk to combine traditional Taiwanese puppetry, the Internet of Things (IoT), and 3D avatar models. AvatarTalk enables puppeteers to control avatar puppets on the web using either motion capture gloves or camera-based image recognition. This transformative performance approach not only facilitates remote performances and multi-screen presentations across various devices but also introduces a fresh gesture interpretation technique for dance artists. AvatarTalk supports a mechanism that can accommodate new control and puppet devices from other approaches through the IoT-based microservice concept. We develop a calibration procedure to enable the accurate capture of hand gestures to manipulate the movements of virtual puppets, empowering them to perform fundamental traditional puppetry poses such as nodding, bowing, and synchronized hand movements. We have conducted experiments to show the accuracy of AvatarTalk calibration. Our study indicates that AvatarTalk can almost detect the right gestures (98.75%-100% recall) and very seldom mistake the wrong gestures (90.8%-100% precision). Additionally, we also provide the mechanism to measure the delays of controlling the puppets. An analytic model is proposed to design the delay times of the messages between the control device of a puppeteer and the AvatarTalk server. In the current AvatarTalk implementation, if the message delay does not exceed 0.1 seconds, four puppeteers can synchronize their actions if the elapsed time between two actions of a puppeteer is longer than 0.3 second.

show abstract

IoT-Based Strawberry Disease Detection With Wall-Mounted Monitoring Cameras

Cited by 5 publications

References 25 publications

ViolinTalk: Violin Robots as Internet of Things Devices

ViolinTalk: Violin Robots as Internet of Things Devices

FenceTalk: Exploring False Negatives in Moving Object Detection

An IoT-Based Microservice Platform for Virtual Puppetry Performance

Contact Info

Product

Resources

About