Deep learning of binary hash codes for fast image retrieval

By Kevin Qinghong Lin, Huei-Fang Yang, Jen-Hao Hsiao, Chu-Song Chen. 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2015

607 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/cf8b46f6882ab889ef3a765bccd0d3e00a9e1b41

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Yi Yang
2. Tao A. Xiang
3. Dacheng Tao
4. Luc Van Gool
5. Xiaogang Wang
6. Qi Tian
7. Bernt Schiele
8. Ming-Hsuan Yang

Citations in the corpus

• SIFT Meets CNN: A Decade Survey of Instance Retrieval (2016, methodology): “These methods take the original image as input and produce a learned feature before binarization [159], [160].”
• A Modulation Module for Multi-task Learning with Applications in Image Retrieval (2018, result): “We do not study the efficiency problem in image retrieval as in many prior works [23,26,24,25].”
• Learning Compact Binary Descriptors with Unsupervised Deep Neural Networks (2016, background): “Among the recent studies of deep learning and binary codes learning, Xia et al. [47] and Lai et al. [21] take deep CNN to learn a set of hash functions, but they require pair-wised similarity labels or triplets training data.”
• Medical image retrieval using deep convolutional neural network (2017, methodology): “Deep learning methods have been applied to CBIR task in recent studies [8, 17, 18], but there is less attention on exploring deep learning methods for CBMIR task.”
• Variational Deep Semantic Hashing for Text Documents (2017, background): “…of images. typical deep learning architectures for hash function learning consist of CNNs layers for representation learning and hash function layers which then transform the representation to supervisory signals. loss functions could be pointwise [19], pairwise [4], or listwise [16].”
• Egocentric Shopping Cart Localization (2018, methodology): “Other methods represent images by extracting the activations of an intermediate layer of a model trained for classification/regression on the target dataset [15].”
• A Decade Survey of Content Based Image Retrieval Using Deep Learning (2020, methodology): “other supervised approaches are deep binary hash codes [56], deep hashing network [73], deep spherical quantization [97], adaptive loss based supervised deep learning to hash [129], etc.”
• Deep Supervised Discrete Hashing (2017, methodology): “of previous deep hashing methods (e.g., the semantic information is not fully exploited). Recent works try to divide the whole learning process into two streams under the multitask learning framework [11, 21, 22]. The hash stream is used to learn the hash function, while the classification stream is ”
• Wafer Map Defect Pattern Classification and Image Retrieval Using Convolutional Neural Network (2018, methodology): “To achieve this goal, we follow the similar approach described in [9].”
• Towards Robust Deep Neural Networks with BANG (2016, background): “Machine learning is broadly used in various real-world vision applications and recent advances in deep learning have made deep neural networks the most powerful learning models that can be successfully applied to different vision problems [27, 25, 7, 28, 20, 14, 18, 17, 29].”
• Beyond One-hot Encoding: lower dimensional target embedding (2018, methodology): “In addition, for approaches learning codes as latent variables, to find the optimal ones in terms of class separability or error correction is not guaranteed [29].”
• Deep Learning for Free-Hand Sketch: A Survey (2020, background): “pipeline has been widely studied in photo-oriented deep hashing [237], [238], where the hashing layer is typically fullyconnected with sigmoid or tanh activation, and a discrete binary constraint.”
• SketchMate: Deep Hashing for Million-Scale Human Sketch Retrieval (2018, methodology): “Recently, deep hashing learning has shown superiority on better preserving the semantic information when compared with shallow methods [28, 14, 21].”
• Feature Learning Based Deep Supervised Hashing with Pairwise Labels (2015, methodology): “Representative pointwise label based methods include CCA-ITQ [3], supervised discrete hashing (SDH) [25] and the deep hashing method in [15].”