On top of previous work in automatic violin transcription we present a set of straight forward low level descriptors for assisting the transcription techniques and saving computational cost. Proposed descriptors have been tested against a database of 1500 violin notes and double stops. Keywords

semantic categories in the perception of image similarity. We then used these data to discover low-level features that best describe each category. Finally, we devised an image similarity metric that embodies our findings and models the behavior of subjects in categorizing images and measuring

Old printed photos are affected by several typical damages, due to age and bad preservation. “Foxing ” defects look like red-brownish spots onto the paper of the printed photo. Similar features can be seen in the digitized copies. In this paper we propose a set of low level descriptors to extract

speech adapted pattern recognition framework for measuring

In this paper we combine previous works done in the domain of image processing used to extract automatically low-level descriptors and previous works in the eld of data mining used to find a kind of dependencies between variables called association rules. Using

an anti-spoofing solution based on a set of low-level feature descriptors capable of distinguishing between ‘live ’ and ‘spoof ’ images and videos. The proposed method explores both spatial and temporal information to learn distinctive characteristics between the two classes. Experiments conducted

In this paper, we propose a spoken sentence retrieval system based on MPEG-7 audio LLDs (Low-Level Descriptors). Our system retrieves the spoken sentence by a two-steps sentence matching method. First, we locate several possible segments that are similar with the user’s query in spoken documents

basis of the image representation. Regions of the BPT are described by low-level descriptors. Finally, relevance feedback is used to avoid the need of manually setting the weights associated to each descriptor. 1

were pooled together and grouped into 12 low level descriptor types and 6 functional types. For each of these groups, classification results using Support Vector Machines and Random Forests are reported for the full set of features, and for 150 features each with the highest individual Information Gain

Abstract — Affective Computing has grown an important field in today’s man-machine-interaction, and the acoustic speech signal is very popular as basis for an automatic classification at the moment. However, recognition performances reported today are mostly not sufficient for a real usage within working systems. Therefore we want to improve on this challenge by evolutionary programming. As a starting point we use prosodic, voice quality and articulatory feature contours. We next propose systematic derivation of functionals by means of descriptive statistics. In order to analyze cross-feature information and feature permutations we use Genetic Algorithms, as a complete coverage of possible alterations is NP-hard. The final attribute set is at the same time optimized by reduction to the most relevant information in order to reduce complexity for the classifier and ensure real-time capability during extraction process. Classification is fulfilled by diverse machine learning methods for utmost discrimination power. We decided for two public databases, namely the Berlin Emotional Speech Database, and the Danish Emotional Speech Corpus for test-runs. These clearly show the high effectiveness of the suggested approach.