Compact Representation of High-Dimensional Feature Vectors for Large-Scale Image Recognition and Retrieval

Feature Selection Based on Intrusive Outliers Rather Than All Instances

Feature selection (FS) has recently attracted considerable attention in many fields. Highly-overlapping classes and skewed distributions of data within classes have been found in various classification tasks. Most existing FS methods are all instance-based, which ignores the significant differences in characteristics between the particular outliers and the main body of the class, causing confusion for classifiers. In this paper, we propose a novel supervised FS method, Intrusive Outliers-based Feature Selection (IOFS), to find out what kind of outliers lead to misclassification and exploit the characteristics of such outliers. In order to accurately identify the intrusive outliers (IOs), we provide a density-mean center algorithm to obtain the appropriate representative of a class. A special distance threshold is given to obtain the candidate for IOs. Combining with several metrics, mathematical formulations are provided to evaluate the overlapping degree of the intrusive class pairs. Features with high overlapping degrees are assigned to low rankings in IOFS method. An extension of IOFS based on a small number of extreme IOs, called E-IOFS, is also proposed. Three theoretical proofs are provided for the essential theoretical basis of IOFS. Experiments comparing against various state-of-the-art methods on eleven benchmark datasets show that IOFS is rational and effective, especially on the datasets with higher overlapping classes. And E-IOFS almost always outperforms IOFS.

Evaluation of scientific and technological achievements in colleges based on machine learning and image feature retrieval

At present, experts and scholars have conducted more research on the ability of colleges and universities to transform scientific and technological achievements. However, they pay more attention to the holistic research on the transformation of scientific and technological achievements in colleges and universities across the country, while rarely divide the research objects in detail. In order to improve the evaluation effect of scientific and technological achievements in colleges and universities, this paper builds a university science and technology achievement evaluation system based on machine learning and image feature retrieval on the basis of analyzing the needs of high-tech achievement evaluation. The system has certain flexibility. Moreover, this study selects the appropriate network architecture based on the actual data and mission objectives of the high-tech achievement evaluation. In addition, this paper proposes a FT-GRU model of a gated recurrent unit network incorporating N nearest neighbor text, and a more stable model structure is obtained through system optimization. Finally, this study designs experiments to verify the performance of the model. The research results show that the university science and technology achievement evaluation system based on machine learning and image feature retrieval constructed in this study meets the expected goals and has certain practical significance.

Action Recognition in Still Images With Minimum Annotation Efforts

We focus on the problem of still image-based human action recognition, which essentially involves making prediction by analyzing human poses and their interaction with objects in the scene. Besides image-level action labels (e.g., riding, phoning), during both training and testing stages, existing works usually require additional input of human bounding boxes to facilitate the characterization of the underlying human-object interactions. We argue that this additional input requirement might severely discourage potential applications and is not very necessary. To this end, a systematic approach was developed in this paper to address this challenging problem of minimum annotation efforts, i.e., to perform recognition in the presence of only image-level action labels in the training stage. Experimental results on three benchmark data sets demonstrate that compared with the state-of-the-art methods that have privileged access to additional human bounding-box annotations, our approach achieves comparable or even superior recognition accuracy using only action annotations in training. Interestingly, as a by-product in many cases, our approach is able to segment out the precise regions of underlying human-object interactions.