Feature hashing

About: Feature hashing is a research topic. Over the lifetime, 993 publications have been published within this topic receiving 51462 citations.

Learning Discriminative Feature Representations for Visual Categorization

Abstract: Learning discriminative feature representations has attracted a great deal of attention due to its potential value and wide usage in a variety of areas, such as image/video recognition and retrieval, human activities analysis, intelligent surveillance and human-computer interaction. In this thesis we first introduce a new boosted key-frame selection scheme for action recognition. Specifically, we propose to select a subset of key poses for the representation of each action via AdaBoost and a new classifier, namely WLNBNN, is then developed for final classification. The experimental results of the proposed method are 0.6% - 13.2% better than previous work. After that, a domain-adaptive learning approach based on multiobjective genetic programming (MOGP) has been developed for image classification. In this method, a set of primitive 2-D operators are randomly combined to construct feature descriptors through the MOGP evolving and then evaluated by two objective fitness criteria, i.e., the classification error and the tree complexity. Later, the (near-)optimal feature descriptor can be obtained. The proposed approach can achieve 0.9% ∼ 25.9% better performance compared with state-of-the-art methods. Moreover, effective dimensionality reduction algorithms have also been widely used for obtaining better representations. In this thesis, we have proposed a novel linear unsupervised algorithm, termed Discriminative Partition Sparsity Analysis (DPSA), explicitly considering different probabilistic distributions that exist over the data points, simultaneously preserving the natural locality relationship among the data. All these above methods have been systematically evaluated on several public datasets, showing their accurate and robust performance (0.44% - 6.69% better than the previous) for action and image categorization. Targeting efficient image classification , we also introduce a novel unsupervised framework termed evolutionary compact embedding (ECE) which can automatically learn the task-specific binary hash codes. It is regarded as an optimization algorithm which combines the genetic programming (GP) and a boosting trick. The experimental results manifest ECE significantly outperform others by 1.58% - 2.19% for classification tasks. In addition, a supervised framework, bilinear local feature hashing (BLFH), has also been proposed to learn highly discriminative binary codes on the local descriptors for large-scale image similarity search. We address it as a nonconvex optimization problem to seek orthogonal projection matrices for hashing, which can successfully preserve the pairwise similarity between different local features and simultaneously take image-to-class (I2C) distances into consideration. BLFH produces outstanding results (0.017% - 0.149% better) compared to the state-of-the-art hashing techniques.

Modified geometric hashing for face database indexing

Abstract: This paper presents a modified geometric hashing technique to index the database of facial images. The technique makes use of minimum amount of search space and memory to provide best matches with high accuracy against a query image. Features are extracted using Speeded-Up Robust Features (SURF) operator. To make these features invariant to translation, rotation and scaling, a pre-processing technique consisting of mean centering, principal components, rotation and normalization has been proposed. The proposed geometric hashing is used to hash these features to index each facial image in the database. It has achieved more than 99% hit rate for top 4 best matches.

Perceptual Hashing of Deep Convolutional Neural Networks for Model Copy Detection

Abstract: In recent years, many model intellectual property (IP) proof methods for IP protection have been proposed, such as model watermarking and model fingerprinting. However, with the increasing number of models transmitted and deployed on the Internet, quickly finding the suspect model among thousands of models on model-sharing platforms such as GitHub is in great demand, which concurrently triggers the new security problem of model copy detection for IP protection. As an important part of the model IP protection system, the model copy detection task has not received enough attention. Due to the high computational complexity, both model watermarking and model fingerprinting lack the capability to efficiently find suspected infringing models among tens of millions of models. In this article, inspired by the hash-based image retrieval methods, we introduce a novel model copy detection mechanism: perceptual hashing for convolutional neural networks (CNNs). The proposed perceptual hashing algorithm can convert the weights of CNNs to fixed-length binary hash codes so that the lightly modified version has the similar hash code as the original model. By comparing the similarity of a pair of hash codes between a query model and a test model in the model library, similar versions of a query model can be retrieved efficiently. To the best of our knowledge, this is the first perceptual hashing algorithm for deep neural network models. Specifically, we first select the important model weights based on the model compression theory, then calculate the normal test statistics (NTS) on the segments of important weights, and finally encode the NTS features into hash codes. The experiment performed on a model library containing 3,565 models indicates that our perceptual hashing scheme has a superior copy detection performance.

Enhanced chained and Cuckoo hashing methods for multi-core CPUs

Abstract: A hash table is a fundamental data structure implementing an associative memory that maps a key to its associative value Besides, the paradigm of micro-architecture design of CPUs is shifting away from faster uniprocessors toward slower chip multiprocessors In this paper, we propose enhanced chained hashing and Cuckoo hashing methods for modern computers having a lot of CPU cores with exploiting CPU cache line and hardware level lock-free operations The proposed methods outperform the existing methods in most cases and are very scalable in terms of the number of CPU cores In addition, their performances do not degrade much even with a high fill factor (eg, 90 %) Through extensive experiments using Intel 32-core machine, we have shown our proposed methods improve performance compared with the state-of-the-art version of the four exiting major hashing methods of linear, chained, Cuckoo, and Hopscotch

Multiple kernel learning for image indexing

Abstract: The paper presents a novel framework for learning the hash functions for indexing through Multiple Kernel Learning. The Distance Based Hashing function is applied which does the object projection to hash space by preserving inter object distances. In recent works, the kernel matrix has been proved to be more accurate representation of similarity in various recognition problems. Our framework learns the optimal kernel for hashing by parametrized linear combination of base kernels. A novel application of Genetic Algorithm for the optimization of kernel combination parameters is presented. We also define new texture based feature representation for images. Our proposed framework can also be applied for optimal combination of multiple sources for indexing. The evaluation of the proposed framework is presented for CIFAR-10 dataset by applying individual and combination of different features. Additionally, the primary experimental results with MNIST dataset is also presented.