Music information retrieval (MIR) is an emerging research area that receives growing attention from both the community and music industry. It addresses problem of querying retrieving certain types large data set. Classification a fundamental in MIR. Many tasks MIR can be naturally cast classification setting, such as genre classification, mood artist recognition, instrument etc. annotation, new has attracted much recent years, also general sense. Due to importance research, rapid development...

Many contour-based image corner detectors are based on the curvature scale-space (CSS).We identify weaknesses of CSS-based detectors.First, "curvature" itself by its "definition" is very much sensitive to local variation and noise curve, unless an appropriate smoothing carried out beforehand.In addition, calculation involves derivatives up second order, which may cause instability errors in result.Second, Gaussian causes changes curve it difficult select smoothing-scale, resulting poor...

Feature extraction is a key issue in content-based image retrieval (CBIR). In the past, number of texture features have been proposed literature, including statistic methods and spectral methods. However, most them are not able to accurately capture edge information which important feature an image. Recent researches on multi-scale analysis, especially curvelet research, provide good opportunity extract more accurate for retrieval. Curvelet was originally denoising has shown promising...

The development of robust and accurate methods for automatic building detection regularisation using multisource data continues to be a challenge due point cloud sparsity, high spectral variability, urban objects differences, surrounding complexity, misalignment. To address these challenges, constraints on object’s size, height, area, orientation are generally benefited which adversely affect the performance. Often buildings either small in under shadows or partly occluded ousted during...

Shape description is one of the key parts image content in MPEG-7. Most existing shape descriptors are usually either application dependent or non-robust, making them undesirable for generic description. A Fourier descriptor (GFD) proposed to overcome drawbacks representation techniques. The derived by applying a 2D transform on polar raster sampled image. acquired independent and robust. Experimental results show GFD outperforms Zernike moment (ZMD) which has been as

Similarity measurement is one of the key issues in content based image retrieval (CBIR). In CBIR, images are represented as features database. Once extracted from indexed images, becomes similarity between features. Many measurements exist. A number commonly used described and evaluated this paper. They a standard shape Results show that city block distance /spl chi//sup 2/ Statistics measure outperform other terms both accuracy efficiency.

Corner detectors have many applications in computer vision and image identification retrieval. Contour-based corner directly or indirectly estimate a significance measure (e.g., curvature) on the points of planar curve, select curvature extrema as corners. While an extensive number contour-based been proposed over last four decades, there is no comparative study recently detectors. This paper attempt to fill this gap. The general framework detection presented, two major issues-curve...

Domain adaptation aims to transfer knowledge from a domain with adequate labeled samples scarce samples. Prior research has introduced various open set settings in the literature extend applications of methods real-world scenarios. This paper focuses on type setting where target both private ('unknown classes') label space and shared ('known space. However, source only 'known classes' Prevalent distribution-matching are inadequate such that demands smaller larger diverse more classes. For...

Conventional image classification methods usually require a large number of training samples for the model. However, in practical scenarios, amount available sample data is often insufficient, which easily leads to overfitting network construction. Few-shot learning provides an effective solution this problem and has been hot research topic. This paper intensive survey on state-of-the-art techniques based few-shot learning. According different deep mechanisms, existing algorithms are divided...

Shape representation is a fundamental issue in the newly emerging multimedia applications. In content based image retrieval (CBIR), shape an important low level feature. Many representations have been proposed. However, for CBIR, should satisfy several properties such as affine invariance, robustness, compactness, computation complexity and perceptual similarity measurement. Against these properties, this paper we attempt to study compare descriptors which widely adopted they are: Fourier...

Gabor filter is widely used to extract texture features from images for image retrieval. A number of parameters (number scales and orientations mask size) are in the filter. In reported work so far, these seem be chosen without proper explanations. this paper, we investigate effects different on