Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification
This paper presents an extensive research carried out for enhancing the performances of convolutional neural network (CNN) object detectors applied to municipal waste identification. In order to obtain an accurate and fast CNN architecture, several types of Single Shot Detectors (SSD) and Regional P...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-10-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/10/20/7301 |
| _version_ | 1797550551643193344 |
|---|---|
| author | Daniel Octavian Melinte Ana-Maria Travediu Dan N. Dumitriu |
| author_facet | Daniel Octavian Melinte Ana-Maria Travediu Dan N. Dumitriu |
| author_sort | Daniel Octavian Melinte |
| collection | DOAJ |
| description | This paper presents an extensive research carried out for enhancing the performances of convolutional neural network (CNN) object detectors applied to municipal waste identification. In order to obtain an accurate and fast CNN architecture, several types of Single Shot Detectors (SSD) and Regional Proposal Networks (RPN) have been fine-tuned on the TrashNet database. The network with the best performances is executed on one autonomous robot system, which is able to collect detected waste from the ground based on the CNN feedback. For this type of application, a precise identification of municipal waste objects is very important. In order to develop a straightforward pipeline for waste detection, the paper focuses on boosting the performance of pre-trained CNN Object Detectors, in terms of precision, generalization, and detection speed, using different loss optimization methods, database augmentation, and asynchronous threading at inference time. The pipeline consists of data augmentation at the training time followed by CNN feature extraction and box predictor modules for localization and classification at different feature map sizes. The trained model is generated for inference afterwards. The experiments revealed better performances than all other Object Detectors trained on TrashNet or other garbage datasets with a precision of 97.63% accuracy for SSD and 95.76% accuracy for Faster R-CNN, respectively. In order to find the optimal higher and lower bounds of our learning rate where the network is actually learning, we trained our model for several epochs, updating the learning rate after each epoch, starting from 1 × 10<sup>−10</sup> and decreasing it until reaching 1 × 10<sup>−1</sup>. |
| first_indexed | 2024-03-10T15:30:58Z |
| format | Article |
| id | doaj.art-d772b52acc2945d1bee655f0b97da66e |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T15:30:58Z |
| publishDate | 2020-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-d772b52acc2945d1bee655f0b97da66e2023-11-20T17:38:27ZengMDPI AGApplied Sciences2076-34172020-10-011020730110.3390/app10207301Deep Convolutional Neural Networks Object Detector for Real-Time Waste IdentificationDaniel Octavian Melinte0Ana-Maria Travediu1Dan N. Dumitriu2Institute of Solid Mechanics of the Romanian Academy, 010141 Bucharest, RomaniaInstitute of Solid Mechanics of the Romanian Academy, 010141 Bucharest, RomaniaInstitute of Solid Mechanics of the Romanian Academy, 010141 Bucharest, RomaniaThis paper presents an extensive research carried out for enhancing the performances of convolutional neural network (CNN) object detectors applied to municipal waste identification. In order to obtain an accurate and fast CNN architecture, several types of Single Shot Detectors (SSD) and Regional Proposal Networks (RPN) have been fine-tuned on the TrashNet database. The network with the best performances is executed on one autonomous robot system, which is able to collect detected waste from the ground based on the CNN feedback. For this type of application, a precise identification of municipal waste objects is very important. In order to develop a straightforward pipeline for waste detection, the paper focuses on boosting the performance of pre-trained CNN Object Detectors, in terms of precision, generalization, and detection speed, using different loss optimization methods, database augmentation, and asynchronous threading at inference time. The pipeline consists of data augmentation at the training time followed by CNN feature extraction and box predictor modules for localization and classification at different feature map sizes. The trained model is generated for inference afterwards. The experiments revealed better performances than all other Object Detectors trained on TrashNet or other garbage datasets with a precision of 97.63% accuracy for SSD and 95.76% accuracy for Faster R-CNN, respectively. In order to find the optimal higher and lower bounds of our learning rate where the network is actually learning, we trained our model for several epochs, updating the learning rate after each epoch, starting from 1 × 10<sup>−10</sup> and decreasing it until reaching 1 × 10<sup>−1</sup>.https://www.mdpi.com/2076-3417/10/20/7301artificial intelligencedeep learningreal-time object detectorimage classificationcomputer visionconvolutional neural networks |
| spellingShingle | Daniel Octavian Melinte Ana-Maria Travediu Dan N. Dumitriu Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification Applied Sciences artificial intelligence deep learning real-time object detector image classification computer vision convolutional neural networks |
| title | Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification |
| title_full | Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification |
| title_fullStr | Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification |
| title_full_unstemmed | Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification |
| title_short | Deep Convolutional Neural Networks Object Detector for Real-Time Waste Identification |
| title_sort | deep convolutional neural networks object detector for real time waste identification |
| topic | artificial intelligence deep learning real-time object detector image classification computer vision convolutional neural networks |
| url | https://www.mdpi.com/2076-3417/10/20/7301 |
| work_keys_str_mv | AT danieloctavianmelinte deepconvolutionalneuralnetworksobjectdetectorforrealtimewasteidentification AT anamariatravediu deepconvolutionalneuralnetworksobjectdetectorforrealtimewasteidentification AT danndumitriu deepconvolutionalneuralnetworksobjectdetectorforrealtimewasteidentification |