When listening to foreign radio or TV programs we are able to pick up some information from the way people are interacting with each others and easily identify the most dominant speaker or the person who is interviewed. Our work relies on the existence of clues about speaker roles in acoustic and prosodic low-level features extracted from audio files and from speaker segmentations. In this pape...

Recently, Massive Open Online Courses (MOOCs) have garnered a high level of interest in the media. With larger and larger numbers of students participating in each course, finding useful and informative threads in increasingly crowded course discussion forums becomes a challenging issue for students. In this work, we address this thread overload problem by taking advantage of an adaptive featur...

The prediction of high-level music categories, such as genres, styles, or personal preferences, helps to organise music collections. The relevance of single audio features for automatic classification depends on a certain category. Relevant feature subsets for each classification task can be identified by means of feature selection. Continuing our previous studies on multi-objective feature sel...

Extraction and utilization of high-level semantic features are critical for more effective video retrieval. However, the performance of video retrieval hasn’t benefited much despite of the advances in high-level feature extraction. To make good use of high-level semantic features in video retrieval, we present a method called pointwise mutual information weighted scheme(PMIWS). The method makes...

Feature-based registration has been popular with a variety of features ranging from voxel intensity to Self-Similarity Context (SSC). In this paper, we examine the question on how features learnt using various Deep Learning (DL) frameworks can be used for deformable registration and whether this feature learning is necessary or not. We investigate the use of features learned by different DL met...

Most of the content-based image retrieval systems focus on similarity-based retrieval of images by utilizing color, shape and texture features. For color-based image retrieval, the average color or color-histograms of images are widely used as feature vectors. In this paper, we propose a new searching scheme, called Fuzzy Membership Value-Indexing, to guarantee higher retrieval quality. This sc...

Connectivity analysis methodology is suitable to find representative symptoms of a disease. This methodology describes connections between symptoms in particular way and then chooses the group of symptoms that have the high level of connection, or in other words, have strong interconnections between elements of a group. In this paper we investigate the analogy between connectivity analysis and ...

Low level features like color etc. of an image are really very important for any image retrieval system. This paper implements image classification technique using SURF descriptor and SVM classifier. SURF method which is advanced version of SIFT is used to match feature points of training and test images. SVM classifier based on the outcome of feature points then classifies images. Through the ...

In this paper, we present a transfer learning approach for music classification and regression tasks. We propose to use a pre-trained convnet feature, a concatenated feature vector using the activations of feature maps of multiple layers in a trained convolutional network. We show how this convnet feature can serve as a general-purpose music representation. In the experiments, a convnet is trai...

While there is an enormous amount of music data available, the field of music analysis almost exclusively uses manually designed features. In this work we learn features from music data in a completely unsupervised way and evaluate them on a musical genre classification task. We achieve results very close to state-of-the-art performance which relies on highly hand-tuned feature extractors.