최근 딥러닝은 오디오, 텍스트 및 이미지 데이터와 같은 비 체계적인 데이터를 대상으로 다양한 추정, 분류 및 예측 문제에 사용 및 적용되고 있다. 특히, 의류산업에 적용될 경우 딥러닝 기법을 활용한 의류 인식, 의류 검색, 자동 제품 추천 등의 심층 학습을 기반으로 한 응용이 가능하다. 이 때의 핵심모형은 합성곱 신경망을 사용한 이미지 분류이다. 합성곱 신경망은 입력이 전달되고 출력에 도달하는 과정에서 가중치와 같은 매개 변수를 학습하는 뉴런으로 구성되고, 영상 분류에 가장 적합한 방법론으로 사용된다. 기존의 의류 이미지 분류 작업에서 대부분의 분류 모형은 의류 이미지 자체 또는 전문모델 착용 의류와 같이 통제된 상황에서 촬영되는 온라인 제품 이미지를 사용하여 학습을 수행한다. 하지만 본 연구에서는 통제되지 않은 상황에서 촬영되고 사람들의 움직임과 다양한 포즈가 포함된 스트릿 패션 이미지 또는 런웨이 이미지를 분류하려는 상황을 고려하여 분류 모형을 훈련시키는 효과적인 방법을 제안한다. 이동성을 포착하는 런웨이 의류 이미지로 모형을 학습시킴으로써 분류 모형의 다양한 쿼리 이미지에 대한 적응력을 높일 수 있다. 모형 학습 시 먼저 ImageNet데이터셋을 사용하여 pre-training 과정을 거치고 본 연구를 위해 수집된 32 개 주요 패션 브랜드의 2426개 런웨이 이미지로 구성된 데이터셋을 사용하여 fine-tuning을 수행한다. 학습 과정의 일반화를 고려해 10번의 실험을 수행하고 제안된 모형은 최종 테스트에서 67.2 %의 정확도를 기록했다. 본 연구 모형은 쿼리 이미지가 런웨이 이미지, 제품 이미지 또는 스트릿 패션 이미지가 될 수 있는 다양한 분류 환경에 적용될 수 있다. 구체적으로는 패션 위크에서 모바일 어플리케이션 서비스를 통해 브랜드 검색을 용이하게 하는 서비스를 제공하거나, 패션 잡지사의 편집 작업에 사용되어 브랜드나 스타일을 분류하고 라벨을 붙일 수 있으며, 온라인 쇼핑몰에서 아이템 정보를 제공하거나 유사한 아이템을 추천하는 등의 다양한 목적에 적용될 수 있다.
최근 딥러닝은 오디오, 텍스트 및 이미지 데이터와 같은 비 체계적인 데이터를 대상으로 다양한 추정, 분류 및 예측 문제에 사용 및 적용되고 있다. 특히, 의류산업에 적용될 경우 딥러닝 기법을 활용한 의류 인식, 의류 검색, 자동 제품 추천 등의 심층 학습을 기반으로 한 응용이 가능하다. 이 때의 핵심모형은 합성곱 신경망을 사용한 이미지 분류이다. 합성곱 신경망은 입력이 전달되고 출력에 도달하는 과정에서 가중치와 같은 매개 변수를 학습하는 뉴런으로 구성되고, 영상 분류에 가장 적합한 방법론으로 사용된다. 기존의 의류 이미지 분류 작업에서 대부분의 분류 모형은 의류 이미지 자체 또는 전문모델 착용 의류와 같이 통제된 상황에서 촬영되는 온라인 제품 이미지를 사용하여 학습을 수행한다. 하지만 본 연구에서는 통제되지 않은 상황에서 촬영되고 사람들의 움직임과 다양한 포즈가 포함된 스트릿 패션 이미지 또는 런웨이 이미지를 분류하려는 상황을 고려하여 분류 모형을 훈련시키는 효과적인 방법을 제안한다. 이동성을 포착하는 런웨이 의류 이미지로 모형을 학습시킴으로써 분류 모형의 다양한 쿼리 이미지에 대한 적응력을 높일 수 있다. 모형 학습 시 먼저 ImageNet 데이터셋을 사용하여 pre-training 과정을 거치고 본 연구를 위해 수집된 32 개 주요 패션 브랜드의 2426개 런웨이 이미지로 구성된 데이터셋을 사용하여 fine-tuning을 수행한다. 학습 과정의 일반화를 고려해 10번의 실험을 수행하고 제안된 모형은 최종 테스트에서 67.2 %의 정확도를 기록했다. 본 연구 모형은 쿼리 이미지가 런웨이 이미지, 제품 이미지 또는 스트릿 패션 이미지가 될 수 있는 다양한 분류 환경에 적용될 수 있다. 구체적으로는 패션 위크에서 모바일 어플리케이션 서비스를 통해 브랜드 검색을 용이하게 하는 서비스를 제공하거나, 패션 잡지사의 편집 작업에 사용되어 브랜드나 스타일을 분류하고 라벨을 붙일 수 있으며, 온라인 쇼핑몰에서 아이템 정보를 제공하거나 유사한 아이템을 추천하는 등의 다양한 목적에 적용될 수 있다.
Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, class...
Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image
Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
제안 방법
We compose the training image dataset with runway held in 2017 and collect 2426 images of 32 major fashion brands from Google Image Search with search keyword of brand names. For the training process, we first pre-train the network with large-scales ImageNet dataset, and then fine-tune the network with our collected runway image dataset. The results have shown that our proposed model achieves significant classification accuracy during 10 experiments as the average training accuracy reaches 80.
4%. The architecture consists of 5 convolutional layers, max-pooling layers, dropout layers, and 3 fully-connected layers, which is the first attempt to increase the number of layers and perform large-scaled classification. In 2014, VGGNet becomes the winner with its greater performance of 7.
For pooling layers, 3x3 max-pooling is used to reduce spatial sizes and prevent overfitting. The model uses Rectified Linear Unit (ReLU) activation function after each convolutional layer to improve the nonlinearity.
대상 데이터
However, we proclaim that there was no form of any specific grant from these brand authorities. For each brand, about 75 images are collected. Detailed number of images for each brand is descripted in [Table 3].
It is large-scaled public image dataset widely used in image classification task. It consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. [Table 2] descripts the detailed categorization of ImageNet dataset.
GoogLeNet was able to achieve both improved performance and computational efficiency by using Inception module. It is composed of 22 convolutional layers, 9 Inception modules, which leads to over 100 layers in total with less parameters.
이론/모형
In case of query image, the input image goes through the pre-trained network and then is classified into 32 fashion brand categories. The model is designed based on vanilla GoogLeNet architecture. The learning process includes two stages.
성능/효과
Throughout the experiments, we could reach consistent results. First, [Table 4] shows the overall accuracy and loss, which are 80.6% of train accuracy, 64.1% of validation accuracy, and 67.2% of test accuracy. And if we look at [Table 5], which descripts the average accuracy and loss per step table, the accuracy gets improved along with the time steps and reaches final loss of 1.
For the training process, we first pre-train the network with large-scales ImageNet dataset, and then fine-tune the network with our collected runway image dataset. The results have shown that our proposed model achieves significant classification accuracy during 10 experiments as the average training accuracy reaches 80.6%, 64.1% for validation accuracy, and 67.2% for the final test accuracy.
참고문헌 (20)
Ahn, S., "Deep learning architectures and applications", Journal of Intelligence and Information Systems, Vol.22, No.2(2016), 127-142.
Bossard, L., Dantone, M., Leistner, C., Wengert, C., Quack, T., and Van Gool, L., "Apparel classification with style", Proceedings of Asian Conference on Computer Vision(2012), 321-335.
Eshwar, S. G., Rishikesh, A. V., Charan, N. A., and Umadevi, V., "Apparel classification using convolutional neural networks", Proceedings of IEEE International Conference on ICT in Business Industry & Government(2016), 1-5.
Ferreira, A. and Giraldi, G., "Convolutional neural network approaches to granite tiles classification", Expert Systems with Applications, No.84(2017), 1-11.
Hara, K., Jagadeesh, V., and Piramuthu, R., "Fashion apparel detection: the role of deep convolutional neural network and pose-dependent priors", Proceedings of IEEE Winter Conference on Applications of Computer Vision(2016), 1-9.
Hu, Z., Yan, H., and Lin, X., "Clothing segmentation using foreground and background estimation based on the constrained Delaunay triangulation", Pattern Recognition, Vol. 41, No.5(2008), 1581-1592.
Iliukovich-Strakovskaia, A., Dral, A., and Dral, E., "Using pre-trained models for fine-grained image classification in fashion field", Unpublished results.
Kim, H. J., "Dynamic Hand Gesture Recognition Using CNN Model and FMM Neural Networks", Journal of Intelligence and Information Systems, Vol.16, No.2(2010), 95-108.
Kim, M. and Cho, Y., "A Multimodal Profile Ensemble Approach to Development of Recommender Systems Using Big Data", Journal of Intelligence and Information Systems, Vol.21, No.4(2015), 93-110.
Krizhevsky, A., Sutskever, I., and Hinton, G. E., "Imagenet classification with deep convolutional neural networks", Advances in Neural Information Processing Systems(2012), 1097-1105.
Kwon, O. and Lee, J. S., "Smarter Classification for Imbalanced Data Set and Its Application to Patent Evaluation", Journal of Intelligence and Information Systems, Vol.20, No.1(2014), 15-34.
Liu, S., Song, Z., Liu, G., Xu, C., Lu, H., and Yan, S., "Street-to-shop: Cross-scenario clothing retrieval via parts alignment and auxiliary set", Proceedings of IEEE Computer Vision and Pattern Recognition(2012), 3330-3337.
Patki, R. and Suresha, S., "Apparel classification using CNNs", Unpublished results.
Simonyan, K. and Zisserman, A., "Very deep convolutional networks for large-scale image recognition", Unpublished results.
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., and Rabinovich, A., "Going deeper with convolutions", Proceedings of Computer Vision and Pattern Recognition(2015).
Taigman, Y., Yang, M., Ranzato, M. A., and Wolf, L., "Deepface: Closing the gap to human-level performance in face verification", Proceedings of IEEE Conference on Computer Vision and Pattern Recognition(2014), 1701-1708.
Tsantekidis, A., Passalis, N., Tefas, A., Kanniainen, J., Gabbouj, M., and Iosifidis, A., "Forecasting stock prices from the limit order book using convolutional neural networks", Proceedings of Conference on Business Informatics(2017), 7-12.
Wang, N. and Ai, H., "Who blocks who: Simultaneous clothing segmentation for grouping images", Proceedings of IEEE International Conference on Computer Vision(2011), 1535-1542.
Yosinski, J., Clune, J., Bengio, Y., and Lipson, H., "How transferable are features in deep neural networks?", Proceedings of Advances in neural information processing systems(2014), 3320-3328.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.