DSHPoolF: deep supervised hashing based on selective pool feature map for image retrieval
- PDF / 1,528,996 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 88 Downloads / 219 Views
ORIGINAL ARTICLE
DSHPoolF: deep supervised hashing based on selective pool feature map for image retrieval P. Arulmozhi1 · S. Abirami1 Accepted: 6 October 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Deep supervised hashing has turned up to unravel many large-scale image retrieval challenges. Although deep supervised hashing accomplishes good results for image retrieval process, requisite for further improving the retrieval accuracy always remains the primal focus of interest. In Deep hashing methods, feature representation happens at the outset of the fully connected (FC) layers, causing shortage of spatial information owing to its global nature, whereas deeper pooling layers preserve semantically similar information by retaining the images spatial information, which can result in uplifting the retrieval performance. Hereby, for enhancing the image retrieval accuracy through exploring spatial information, a novel way of deep supervised hashing based on Pooled Feature map (DSHPoolF) is proposed to generate compact hash codes that explore the spatial information by weighing the informative Feature maps from the last pooling layer. This is achieved, firstly, by weighing the last pooling layers Feature map in two ways, namely average–max-based pooling and probability-based pooling strategies. Secondly, informative Feature maps are selected with the help of the weights. In addition to this, the informative Feature maps play a key role in optimizing quantization error together with the loss function and classification errors in a single-step, point-wise ranking manner. This proposed DSHPoolF method is assessed using three datasets (CIFAR-10, MNIST and ImageNet) that unveils primitive outcome in comparison with other existing prominent hash-based methods. Keywords Learning-based hashing · Convolutional neural network · Deep supervised hashing · Image retrieval
1 Introduction Recent years witness the explosion of information, where every day millions of images are uploaded in Internet. This causes difficulty in accessing images effectively and efficiently. Hence, content-based image retrieval (CBIR) [2,4,8,23] concept came into existence. It is a method to retrieve similar images for a query image by representing both the database and query images as real-valued features and ranking them by finding the distance among them. This linear matching technique, nowadays, when applied to millions of database images, increases the cost in terms of both time and space.
B
P. Arulmozhi [email protected] S. Abirami [email protected]
1
Department of Information Science and Technology, Anna University, Chennai, Tamilnadu, India
So, in today’s Internet world, dealing with abundant multimedia data encourages the growth of approximate nearest neighbor (ANN) [14,27] concept. It enhances the performance of retrieving images by attempting approximate matching rather than exact match as done in exact neighbor search. Among ANN techniques, hashing [20,26,40] is a well-known ANN model, which maps the im
Data Loading...
