A Comperative Study on Efficacy of CNN VGG-16,  DenseNet121, ResNet50V2, And EfficientNetB0 in Toraja Carving Classification

Herman; An'nisa Pratama Putri; Megat Norulazmi Megat Mohamed Noor; Herdianti Darwis; Lilis Nur Hayati; Irawati; Ihwana As’ad

doi:10.56705/ijodas.v6i1.220

Herman Universitas Muslim Indonesia
An'nisa Pratama Putri Universitas Muslim Indonesia
Megat Norulazmi Megat Mohamed Noor MIIT University Kuala Lumpur
Herdianti Darwis Universitas Muslim Indonesia
Lilis Nur Hayati Universitas Muslim Indonesia
Irawati Universitas Muslim Indonesia
Ihwana As’ad Universitas Muslim Indonesia

DOI: https://doi.org/10.56705/ijodas.v6i1.220

Keywords: Classification, CNN, DenseNet121, EfficientNetB0, Resnet50V2, Toraja Carving, VGG-16

Abstract

Introduction: Passura', or Toraja carvings, are an essential element of the cultural heritage of the Toraja people in Indonesia. These carvings feature complex motifs rooted in nature, folklore, and spiritual symbolism. This study aims to evaluate the efficacy of four Convolutional Neural Network (CNN) architectures—VGG-16, DenseNet121, ResNet50V2, and EfficientNetB0—in classifying seven traditional Toraja carving motifs. Methods: A dataset of 700 images was collected and categorized into seven motif classes. The dataset was split into 80% for training and 20% for validation. Each CNN model was trained for 25 epochs with standard pre-processing, including resizing to 224×224 and normalization. Performance evaluation was conducted based on validation accuracy and confusion matrix analysis to assess classification precision and model overfitting. Results: EfficientNetB0 achieved the highest validation accuracy of 98%, although signs of overfitting were observed. ResNet50V2 followed closely with a validation accuracy of 95.33% and demonstrated the most balanced classification results across all motif categories. VGG-16 and DenseNet121 achieved 94.67% and 81.82%, respectively. Confusion matrix analysis confirmed the robustness of ResNet50V2 in correctly identifying complex patterns. Conclusions: The findings indicate that ResNet50V2 provides a reliable balance between accuracy and generalizability for classifying Toraja carvings, making it suitable for digital preservation of cultural heritage. EfficientNetB0, while achieving higher accuracy, may require additional regularization to avoid overfitting. This study contributes to the development of AI-driven cultural documentation and suggests future research with larger and more diverse datasets to improve model robustness

Downloads

Download data is not yet available.

References

R. Muslimin, “Toraja glyphs: An ethnocomputation study of passura Indigenous icons,” J. Asian Archit. Build. Eng., vol. 16, no. 1, pp. 39–44, 2017, doi: 10.3130/jaabe.16.39.

Y. S. Nugraha, “Ethnomathematical review of Toraja’s typical carving design in geometry transformation learning,” 2019. doi: 10.1088/1742-6596/1280/4/042020.

A. Donzelli, “Material words: The aesthetic grammar of Toraja textiles, carvings, and ritual language,” J. Mater. Cult., 2020, doi: 10.1177/1359183519858378.

L. Sudianto, “The Encyclopedia of Virtual Art Carving Toraja - Indonesia,” 2018. doi: 10.1051/matecconf/201816401025.

L. Alzubaidi, “Review of deep learning: concepts, CNN architectures, challenges, applications, future directions,” J. Big Data, vol. 8, no. 1, 2021, doi: 10.1186/s40537-021-00444-8.

S. Minaee, “Image Segmentation Using Deep Learning: A Survey,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 7, pp. 3523– 3542, 2022, doi: 10.1109/TPAMI.2021.3059968.

A. Narin, “Automatic detection of coronavirus disease (COVID-19) using X-ray images and deep convolutional neural networks,” Pattern Anal. Appl., vol. 24, no. 3, pp. 1207–1220, 2021, doi: 10.1007/s10044-021-00984-y.

M. Desai, “An anatomization on breast cancer detection and diagnosis employing multi-layer perceptron neural network (MLP) and Convolutional neural network (CNN),” 2021. doi: 10.1016/j.ceh.2020.11.002.

J. Liu, “Hourly stepwise forecasting for solar irradiance using integrated hybrid models CNN-LSTM-MLP combined with error correction and VMD,” Energy Convers. Manag., vol. 280, 2023, doi: 10.1016/j.enconman.2023.116804.

Y. Gulzar, “Fruit Image Classification Model Based on MobileNetV2 with Deep Transfer Learning Technique,” Sustain., vol. 15, no. 3, 2023, doi: 10.3390/su15031906.

Herman, H. Nasir, M. N. M. M. Noor, T. Hasanuddin, D. Indra, and H. B. Lumentut, “Exploration of CNN Parameters to Measure Performance of LeNet-5 Architecture in Toraja Carving Classification,” in 2024 IEEE 8th International Conference on Signal and Image Processing Applications (ICSIPA), 2024, pp. 1–6. doi: 10.1109/ICSIPA62061.2024.10686353.

N. Wardhani, B. E. W. Asrul, A. R. Tampang, and ..., “Classification of Toraja Wood Carving Motif Images Using Convolutional Neural Network (CNN),” … (Rekayasa Sist. dan …, 2024, doi: 10.29207/resti.v8i4.5897.

M. Caldeira, “Comparison study on convolution neural networks (CNNs) vs. Human visual system (HVS),” 2019. doi: 10.1007/978-3030-19093-4_9.

A. Celeghin, “Convolutional neural networks for vision neuroscience: significance, developments, and outstanding issues,” 2023. doi: 10.3389/fncom.2023.1153572.

R. Nirthika, “Pooling in convolutional neural networks for medical image analysis: a survey and an empirical study,” 2022. doi: 10.1007/s00521-022-06953-8.

H. Darwis, “Max Feature Map CNN with Support Vector Guided SoftMax for Face Recognition,” Int. J. Informatics Vis., vol. 7, no. 3, pp. 959–966, 2023, doi: 10.30630/joiv.7.3.1751.

T. Turay, “Toward Performing Image Classification and Object Detection with Convolutional Neural Networks in Autonomous Driving Systems: A Survey,” IEEE Access, vol. 10, pp. 14076–14119, 2022, doi: 10.1109/ACCESS.2022.3147495.

G. Georgiadis, “Accelerating convolutional neural networks via activation map compression,” 2019. doi: 10.1109/CVPR.2019.00725.

G. Huang, “Convolutional Networks with Dense Connectivity,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 12, pp. 8704–8716, 2022, doi: 10.1109/TPAMI.2019.2918284.

S. A. Albelwi, “Deep Architecture based on DenseNet-121 Model for Weather Image Recognition,” Int. J. Adv. Comput. Sci. Appl., vol. 13, no. 10, pp. 559–565, 2022, doi: 10.14569/IJACSA.2022.0131065.

N. Cinar, “A hybrid DenseNet121-UNet model for brain tumor segmentation from MR Images,” Biomed. Signal Process. Control, vol. 76, 2022, doi: 10.1016/j.bspc.2022.103647.

M. Rahimzadeh, “A modified deep convolutional neural network for detecting COVID-19 and pneumonia from chest X-ray images based on the concatenation of Xception and ResNet50V2,” Informatics Med. Unlocked, vol. 19, 2020, doi: 10.1016/j.imu.2020.100360.

A. K. Das, “Automatic COVID-19 detection from X-ray images using ensemble learning with convolutional neural network,” Pattern Anal. Appl., vol. 24, no. 3, pp. 1111–1124, 2021, doi: 10.1007/s10044-021-00970-4.

A. Xiao, “Intraoperative Glioma Grading Using Neural Architecture Search and Multi-Modal Imaging,” IEEE Trans. Med. Imaging, vol. 41, no. 10, pp. 2570–2581, 2022, doi: 10.1109/TMI.2022.3166129.

M. A. Joodi, “Increasing validation accuracy of a face mask detection by new deep learning model-based classification,” Indones. J. Electr. Eng. Comput. Sci., vol. 29, no. 1, pp. 304–314, 2023, doi: 10.11591/ijeecs.v29.i1.pp304-314.

Y. I. Sulistya, E. T. Br Bangun, and D. A. Tyas, “CNN Ensemble Learning Method for Transfer learning: A Review,” Ilk. J. Ilm., vol. 15, no. 1, pp. 45–63, 2023, doi : 10.33096/ilkom.v15i1.1541.45-63