Expert role in image classification using CNN for hard to identify object: distinguishing batik and its imitation

Abstract

<span id="docs-internal-guid-25a2977b-7fff-96bd-b93a-19bd55e68ea7"><span>In this research we try to solve the recognition problem in differentiating between batik and its imitation. Batik is an Indonesian heritage of process in making traditional textile product that is now endangered by the existence of imitation products. We try to compare two popular CNN model to classify batik products into five classes. The classes are tulis, cap, print warna, print malam, cabut warna. Tulis and cap are genuine batik, and the other three are an imitation. We realize that this problem is go beyond the recognition of fine grained image problem, it is a hard to identify image problem because even the batik experts is having a hard time identifying batik and its imitation if only based on its picture. The two CNN models, inceptionV3 and mobilenetV2 were trained on three types of image. One type is a freely taken image, the other two were taken based on the experts suggestion. The accuracy score shows that the model trained with the suggestion based picture perform better than the one trained with the random picture.</span></span>

Modernizing Batik Clothes for Community Enterprises Using Creative Design and Colorimetry

“…Nevertheless, a computer-aided design was not employed in this work to avoid overloading them with too many new things at once. Computer programs were previously proven successful in batik designs [17][18][19][20], but the local batik artisans were not keen to use them. The gradual incorporation of new techniques into traditional craftsmanship is an appropriate way to transform batik into modern fashion without losing its local identity and charm.…”

“…Laitupa et al demonstrated the computer-based creation of new batik motifs [17], whereas Tian et al developed automatic batik flower patterns based on fractal geometry [18]. For pattern recognition, neural networks and deep learning have been implemented on batik fabrics [19,20]. To characterize fabric colors, smartphone colorimetry was demonstrated and compared with the standard CIE L*a*b* color measurement using a HunterLab spectrophotometer [21].…”

Batik Images Retrieval Using Pre-trained model and K-Nearest Neighbor

“…The structure of MobileNetV2 can be seen in Fig 3 . MobileNetV2 has been tested on ImageNet classification. In addition, MobileNetV2 was also tested on pattern recognition of traditional clothes [24] and batik image classification [25]. The MobileNetV2 model uses a convolution block with a unique property that separates the model network's expressiveness capacity by using an input bottleneck [26].…”

“…The MobileNetV2 model uses a convolution block with a unique property that separates the model network's expressiveness capacity by using an input bottleneck [26]. 2) InceptionV3 and InceptionResNetV2: InceptionV3 was introduced by Widyantoko et al [25]. InceptionV3 enhances the model network by making the most efficient use of incremental computing with suitable factored convolution and aggressive regularization.…”

“…InceptionV3 enhances the model network by making the most efficient use of incremental computing with suitable factored convolution and aggressive regularization. InceptionV3 is suitable for bigdata classification scenarios [25]. On the other hand, InceptionResNetV2 was introduced by Widyantoko et al [25].…”