Polysemious visual representation based on feature aggregation for large scale image applications

Song, Xinghang; Jiang, Shuqiang; Wang, Shuhui; Li, Liang; Huang, Qingming
January 2015
Multimedia Tools & Applications;Jan2015, Vol. 74 Issue 2, p595
Academic Journal
Multiple image features and multiple semantic concepts from the images have intrinsic and complex relations. These relations influence the effectiveness of image semantic analysis methods, especially on the large scale problems. In this paper, a framework of generating polysemious image representation through three levels of feature aggregation is proposed. In the codebook level aggregation, visual dictionaries are learned for each feature type, and each image feature can be reconstructed with this dictionary. In the semantic level aggregation, the multiple concept distributions are learned with each feature codebook by using the improved local anchor embedding. Then the polysemious representation for for single feature type can be established after this level. In the multiple feature level aggregation, final image polysemious representation is obtained through multiple feature fusion with a weighted pooling approach. Through the proposed framework, multiple feature fusion and multiple semantic descriptions are both achieved in an integrated way. Experimental evaluations on large scale image dataset validate the effectiveness of the proposed method.


Related Articles

  • Structure Representation for Hyper spectral Images Using Binary Classification. Revathi, R. // Language in India;Mar2015, Vol. 15 Issue 3, p175 

    Binary Partition Trees are hierarchical region-based representations of images. They define a reduced set of regions that covers the image support and that spans various levels of resolution. They are attractive for object detection as they tremendously reduce the search space. In this paper,...

  • Convolutional Sparse Coding for Static and Dynamic Images Analysis. Knyazev, B. A.; Chernenkiy, V. M. // Science & Education of Bauman MSTU / Nauka i Obrazovanie of Baum;nov2014, Issue 11, p664 

    The objective of this work is to improve performance of static and dynamic objects recognition. For this purpose a new image representation model and a transformation algorithm are proposed. It is examined and illustrated that limitations of previous methods make it difficult to achieve this...

  • A caricatura política na concepção libertária do periódico A Plebe (1947-1949). Lopes Silva, Zélia // Antíteses;jan-jun2013, Vol. 6 Issue 11, p261 

    This article discusses the meanings of imagistic representations (drawings, cartoons) published in A Plebe (The Plebs), from May 1947 to May 1949, the last period of the Edgard Leuenroth's headship. Created in 1917, the newspaper, supported by libertarian principles, set up as a public sphere...

  • DIGITAL SIGNAGE MEETS BIG DATA. Freedlander, Vern // Sound & Video Contractor;Sep2013, Vol. 31 Issue 9, p10 

    The article reflects on the utility of digital display systems as it can help alleviate communications system within corporate enterprise by application of web and Intranet content, and also informs about the visual communication medium for distribution of live info graphics.

  • 2D Geometry Predicts Perceived Visual Curvature in Context-Free Viewing. Dresp-Langley, Birgitta // Computational Intelligence & Neuroscience;8/5/2015, Vol. 2015, p1 

    Planar geometry was exploited for the computation of symmetric visual curves in the image plane, with consistent variations in local parameters such as sagitta, chordlength, and the curves’ height-to-width ratio, an indicator of the visual area covered by the curve, also called aspect...

  • Image classification using label constrained sparse coding. Liu, Ruijun; Chen, Yi; Zhu, Xiaobin; Hou, Kun // Multimedia Tools & Applications;Dec2016, Vol. 75 Issue 23, p15619 

    Sparse coding has been widely used for feature encoding in recent years. However, the encoded parameters' similarity is ignored with sparse coding. Besides, the label information from which class the local feature is extracted is also ignored. To solve this problem, in this paper, we propose a...

  • Similarity metric learning for sketch-based 3D object retrieval. Furuya, Takahiko; Ohbuchi, Ryutarou // Multimedia Tools & Applications;Dec2015, Vol. 74 Issue 23, p10367 

    Sketch-Based 3D Object Retrieval (SB3DOR) algorithms retrieve 3D models similar to hand-drawn sketch queries. It is one of the most effective modalities to query 3D models by their shape. However, comparison of a sketch, which is a 2D image, with a 3D model is not straightforward. Most of the...

  • Object Bank: An Object-Level Image Representation for High-Level Visual Recognition. Li, Li-Jia; Su, Hao; Lim, Yongwhan; Fei-Fei, Li // International Journal of Computer Vision;Mar2014, Vol. 107 Issue 1, p20 

    It is a remarkable fact that images are related to objects constituting them. In this paper, we propose to represent images by using objects appearing in them. We introduce the novel concept of object bank (OB), a high-level image representation encoding object appearance and spatial location...

  • A top-down event-driven approach for concurrent activity recognition. Voulodimos, Athanasios; Kosmopoulos, Dimitrios; Doulamis, Nikolaos; Varvarigou, Theodora // Multimedia Tools & Applications;Mar2014, Vol. 69 Issue 2, p293 

    In this paper a framework for automatic online workflow recognition in industrial environments where the issue of concurrent activities rises, is presented. The framework consists of three main parts: The first part is devoted to detecting activity in specific Regions of Interest (ROIs) of the...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics