Improving a model of object recognition in images based on a convolutional neural network

Knysh, Bogdan; Kulyk, Yaroslav

doi:10.15587/1729-4061.2021.233786

Cited by 5 publications

(9 citation statements)

References 15 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An important study issue consists in the control of "dead" zones of the activation functions. Two activation functions were used in the proposed model (as in the base one): ReLU (4) and Sofmax (7) [20,21]. Mathematical expressions, descriptions, and features of the application are given in Section 4.…”

Section: Discussion Of the Results Obtained In The Study Of Recognizing The Object Images In Aerial Photographs Using Dcnnmentioning

confidence: 99%

“…The following effectiveness indicators (metrics) were selected as the main indicators characterizing the process of training and estimating the DCNN effectiveness [20]:…”

Section: The Study Materials and Methodsmentioning

confidence: 99%

“…accuracy (precision), that is the ratio of correctly recognized objects to the total number of presumptive and true objects [20]…”

Section: The Study Materials and Methodsmentioning

confidence: 99%

“…N val is the number of images in the test sample; t is the current image; -sensitivity (recall) is the ratio of correctly recognized objects to the total number of objects in the image [20]:…”

Section: The Study Materials and Methodsmentioning

confidence: 99%

“…Analysis of [20,21] has shown that the AlexNet 2012 model has high effectiveness in the automatic classification of objects of various shapes and in various positions.…”

Section: Substantiation Of Architecture and Mathematical Apparatus Used For Implementation Of The Proposed Dcnnmentioning

confidence: 99%

See 4 more Smart Citations

Improvement of the model of object recognition in aero photographs using deep convolutional neural networks

Slyusar

Protsenko

Chernukha

et al. 2021

EEJET

View full text Add to dashboard Cite

Detection and recognition of objects in images is the main problem to be solved by computer vision systems. As part of solving this problem, the model of object recognition in aerial photographs taken from unmanned aerial vehicles has been improved. A study of object recognition in aerial photographs using deep convolutional neural networks has been carried out. Analysis of possible implementations showed that the AlexNet 2012 model (Canada) trained on the ImageNet image set (China) is most suitable for this problem solution. This model was used as a basic one. The object recognition error for this model with the use of the ImageNet test set of images amounted to 15 %. To solve the problem of improving the effectiveness of object recognition in aerial photographs for 10 classes of images, the final fully connected layer was modified by rejection from 1,000 to 10 neurons and additional two-stage training of the resulting model. Additional training was carried out with a set of images prepared from aerial photographs at stage 1 and with a set of VisDrone 2021 (China) images at stage 2. Optimal training parameters were selected: speed (step) (0.0001), number of epochs (100). As a result, a new model under the proposed name of AlexVisDrone was obtained. The effectiveness of the proposed model was checked with a test set of 100 images for each class (the total number of classes was 10). Accuracy and sensitivity were chosen as the main indicators of the model effectiveness. As a result, an increase in recognition accuracy from 7 % (for images from aerial photographs) to 9 % (for the VisDrone 2021 set) was obtained which has indicated that the choice of neural network architecture and training parameters was correct. The use of the proposed model makes it possible to automate the process of object recognition in aerial photographs. In the future, it is advisable to use this model at ground stations of unmanned aerial vehicle complex control when processing aerial photographs taken from unmanned aerial vehicles, in robotic systems, in video surveillance complexes and when designing unmanned vehicle systems

show abstract

Section: Discussion Of the Results Obtained In The Study Of Recognizing The Object Images In Aerial Photographs Using Dcnnmentioning

confidence: 99%

“…The following effectiveness indicators (metrics) were selected as the main indicators characterizing the process of training and estimating the DCNN effectiveness [20]:…”

Section: The Study Materials and Methodsmentioning

confidence: 99%

“…accuracy (precision), that is the ratio of correctly recognized objects to the total number of presumptive and true objects [20]…”

Section: The Study Materials and Methodsmentioning

confidence: 99%

Section: The Study Materials and Methodsmentioning

confidence: 99%

“…Analysis of [20,21] has shown that the AlexNet 2012 model has high effectiveness in the automatic classification of objects of various shapes and in various positions.…”

Section: Substantiation Of Architecture and Mathematical Apparatus Used For Implementation Of The Proposed Dcnnmentioning

confidence: 99%

See 3 more Smart Citations

Improvement of the model of object recognition in aero photographs using deep convolutional neural networks

Slyusar

Protsenko

Chernukha

et al. 2021

EEJET

View full text Add to dashboard Cite

show abstract

Internet Data Risk Model of Sports Events Based on Convolutional Neural Network

Song

2022

2022 European Conference on Natural Language Processing and Information Retrieval (ECNLPIR)

View full text Add to dashboard Cite

Limiting COVID-19 infection by automatic remote face mask monitoring and detection using deep learning with IoT

Alsaydia¹,

Saadallah²,

Malallah³

et al. 2021

EEJET

View full text Add to dashboard Cite

During the current outbreak of the COVID-19 pandemic, controlling and decreasing the possibilities of infections are massively required. One of the most important solutions is to use Artificial Intelligence (AI), which combines both fields of deep learning (DL) and the Internet of Things (IoT). The former one is responsible for detecting any face, which is not wearing a mask. Whereas, the latter is exploited to manage the control for the entire building or a public area such as bus, train station, or airport by connecting a Closed-Circuit Television (CCTV) camera to the room of management. The work is implemented using a Core-i5 CPU workstation attached with a Webcam. Then, MATLAB software is programmed to instruct both Arduino and NodeMCU (Micro-Controller Unit) for remote control as IoT. In terms of deep learning, a 15-layer convolutional neural network is exploited to train 1,376 image samples to generate a reference model to use for comparison. Before deep learning, preprocessing operations for both image enhancement and scaling are applied to each image sample. For the training and testing of the proposed system, the Simulated Masked Face Recognition Dataset ( SMFRD) has been exploited. This dataset is published online. Then, the proposed deep learning system has an average accuracy of up to 98.98 %, where 80 % of the dataset was used for training and 20 % of the samples are dedicated to testing the proposed intelligent system. The IoT system is implemented using Arduino and NodeMCU_TX (for transmitter) and RX (for receiver) for the signal transferring through long distances. Several experiments have been conducted and showed that the results are reasonable and thus the model can be commercially applied

show abstract

Improving a model of object recognition in images based on a convolutional neural network

Cited by 5 publications

References 15 publications

Improvement of the model of object recognition in aero photographs using deep convolutional neural networks

Improvement of the model of object recognition in aero photographs using deep convolutional neural networks

Internet Data Risk Model of Sports Events Based on Convolutional Neural Network

Limiting COVID-19 infection by automatic remote face mask monitoring and detection using deep learning with IoT

Contact Info

Product

Resources

About