Research Article
Image Classification Based on Light Convolutional Neural Network Using Pulse Couple Neural Network
| | Technique | Year | Number of parameters | Dataset | Precision | Recall | F1 | Accuracy (top-1) |
| | Local texture descriptor + SVM [5] | 2017 | — | Caltech-101 | — | — | — | 0.7770 | | CNN applicable in small dataset [8] | 2018 | — | CIFAR-10 | — | — | — | 0.8590 | | Bag of LBP + SVM [6] | 2019 | — | Caltech-101 | 0.6300 | 0.6100 | 0.6100 | 0.7900 | | Standard CNN [38] | 2019 | — | Caltech-256 | 0.95 | 0.96 | 0.9549 | 0.9600 | | Optimization CNN model [9] | 2021 | 2915114 | CIFAR-10 | — | — | — | 0.8240 | | CNN sequential method [10] | 2021 | 289443 | CIFAR-10 | — | High | — | 0.9420 | | ResNet50 [41] | 2021 | — | Caltech-101/Caltech-256/CIFAR-10 | — | — | — | 0.6852/0.8040/0.9079/ | | CNN + DWT [11] | 2022 | — | Caltech-256 | — | — | — | 0.7224 | | Proposed method | 2022 | 10980 | Caltech-101/Caltech-256/CIFAR-10/CIFAR-100 | 0.9722/0.9452/0.9933/0.9547 | 0.9437/0.9494/0.9978/0.9870 | 0.9571/0.9464/0.9955/0.9705 | 0.9270/0.9031/0.9911/0.9438 |
|
|
