Search

Modeling the music mood has wide applications in music categorization, retrieval, and recommendation systems; however, it is challenging to computationally model the affective content of music due to its subjective nature. In this work, a structured regression framework is proposed to model the valence and arousal mood dimensions of music using a single regression model at a linear computational cost. To tackle the subjectivity phenomena, a confidence-interval based estimated consensus is computed by modeling the behavior of various annotators (e.g. biased, adversarial) and is shown to perform better than using the average annotation values. For a compact feature representation of music clips, variational Bayesian inference is used to learn the Gaussian mixture model representation of acoustic features and chord-related features are used to improve the valence estimation by probing the chord progressions between chroma frames. The dimensionality of features is further reduced using an adaptive version of kernel PCA. Using an efficient implementation of twin Gaussian process for structured regression, the proposed work achieves a significant improvement in R2 for arousal and valence dimensions relative to state-of-the-art techniques on two benchmark datasets for music mood estimation. [ABSTRACT FROM AUTHOR]

Annotations are critical for machine learning and developing computer aided diagnosis (CAD) algorithms. Good performance of CAD is critical to their adoption, which generally rely on training with a wide variety of annotated data. However, a vast amount of medical data is either unlabeled or annotated only at the image-level. This poses a problem for exploring data driven approaches like deep learning for CAD. In this paper, we propose a novel crowdsourcing and synthetic image generation for training deep neural net-based lesion detection. The noisy nature of crowdsourced annotations is overcome by assigning a reliability factor for crowd subjects based on their performance and requiring region of interest markings from the crowd. A generative adversarial network-based solution is proposed to generate synthetic images with lesions to control the overall severity level of the disease. We demonstrate the reliability of the crowdsourced annotations and synthetic images by presenting a solution for training the deep neural network (DNN) with data drawn from a heterogeneous mixture of annotations. Experimental results obtained for hard exudate detection from retinal images show that training with refined crowdsourced data/synthetic images is effective as detection performance in terms of sensitivity improves by 25%/27% over training with just expert-markings.

Soft biometrics enable human description and identification from low‐quality surveillance footage. This study premises the design, collection and analysis of a novel crowdsourced dataset of comparative soft biometric body annotations, obtained from a richly diverse set of human annotators. The authors annotate 100 subject images to provide a coherent, in‐depth appraisal of the collected annotations and inferred relative labels. The dataset includes gender as a comparative trait and the authors find that comparative labels characteristically contain additional discriminative information over traditional categorical annotations. Using the authors' pragmatic dataset, semantic recognition is performed by inferring relative biometric signatures using a RankSVM algorithm. This demonstrates a practical scenario, reproducing responses from a video surveillance operator searching for an individual. The approach can reliably return the correct match in the top 7% of results with ten comparisons, or top 13% of results using just five sets of subject comparisons. [ABSTRACT FROM AUTHOR]