FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing
Many heavy and lightweight convolutional neural networks (CNNs) require large datasets and parameter tuning. Moreover, they consume time and computer resources. A new lightweight model called FlexibleNet was created to overcome these obstacles. The new lightweight model is a CNN scaling-based model...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/1/272 |
| _version_ | 1797431221612969984 |
|---|---|
| author | Mohamad M. Awad |
| author_facet | Mohamad M. Awad |
| author_sort | Mohamad M. Awad |
| collection | DOAJ |
| description | Many heavy and lightweight convolutional neural networks (CNNs) require large datasets and parameter tuning. Moreover, they consume time and computer resources. A new lightweight model called FlexibleNet was created to overcome these obstacles. The new lightweight model is a CNN scaling-based model (width, depth, and resolution). Unlike the conventional practice, which arbitrarily scales these factors, FlexibleNet uniformly scales the network width, depth, and resolution with a set of fixed scaling coefficients. The new model was tested by qualitatively estimating sequestered carbon in the aboveground forest biomass from Sentinel-2 images. We also created three different sizes of training datasets. The new training datasets consisted of six qualitative categories (no carbon, very low, low, medium, high, and very high). The results showed that FlexibleNet was better or comparable to the other lightweight or heavy CNN models concerning the number of parameters and time requirements. Moreover, FlexibleNet had the highest accuracy compared to these CNN models. Finally, the FlexibleNet model showed robustness and low parameter tuning requirements when a small dataset was provided for training compared to other models. |
| first_indexed | 2024-03-09T09:41:47Z |
| format | Article |
| id | doaj.art-8170433b73c24bc0aa0018ac12af0917 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T09:41:47Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-8170433b73c24bc0aa0018ac12af09172023-12-02T00:52:17ZengMDPI AGRemote Sensing2072-42922023-01-0115127210.3390/rs15010272FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote SensingMohamad M. Awad0Remote Sesning Center, National Council for Scientific Research, Beirut 11072260, LebanonMany heavy and lightweight convolutional neural networks (CNNs) require large datasets and parameter tuning. Moreover, they consume time and computer resources. A new lightweight model called FlexibleNet was created to overcome these obstacles. The new lightweight model is a CNN scaling-based model (width, depth, and resolution). Unlike the conventional practice, which arbitrarily scales these factors, FlexibleNet uniformly scales the network width, depth, and resolution with a set of fixed scaling coefficients. The new model was tested by qualitatively estimating sequestered carbon in the aboveground forest biomass from Sentinel-2 images. We also created three different sizes of training datasets. The new training datasets consisted of six qualitative categories (no carbon, very low, low, medium, high, and very high). The results showed that FlexibleNet was better or comparable to the other lightweight or heavy CNN models concerning the number of parameters and time requirements. Moreover, FlexibleNet had the highest accuracy compared to these CNN models. Finally, the FlexibleNet model showed robustness and low parameter tuning requirements when a small dataset was provided for training compared to other models.https://www.mdpi.com/2072-4292/15/1/272peri-urban forestslightweight convolutional neural networkFlexibleNetcarbon sequestrationremote sensing |
| spellingShingle | Mohamad M. Awad FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing Remote Sensing peri-urban forests lightweight convolutional neural network FlexibleNet carbon sequestration remote sensing |
| title | FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing |
| title_full | FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing |
| title_fullStr | FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing |
| title_full_unstemmed | FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing |
| title_short | FlexibleNet: A New Lightweight Convolutional Neural Network Model for Estimating Carbon Sequestration Qualitatively Using Remote Sensing |
| title_sort | flexiblenet a new lightweight convolutional neural network model for estimating carbon sequestration qualitatively using remote sensing |
| topic | peri-urban forests lightweight convolutional neural network FlexibleNet carbon sequestration remote sensing |
| url | https://www.mdpi.com/2072-4292/15/1/272 |
| work_keys_str_mv | AT mohamadmawad flexiblenetanewlightweightconvolutionalneuralnetworkmodelforestimatingcarbonsequestrationqualitativelyusingremotesensing |