Searching music by emotion has always been strongly needed by users. Since music streaming applications usually have tens millions of music pieces in database, it is impossible to label emotion tags for each music piece manually. It is desired that an intelligent algorithm can recognize emotion expressed by music automatically. Valence-Arousal model is a widely used for representing emotion, but the angle of vectors on V-A plane labeled by different raters usually varies greatly, which makes it difficult to train classification models. We are trying to introduce a label space defined by pairs of antonyms, which makes emotion label relatively objective. We also used a variant model of recurrent neural network in the paper, in this model, RNN is a mean to extract features from melody, and with other features calculated by normal machine learning algorithms, this model can make a good prediction of emotions.
ABSTRACT:Multispectral Lidar System can emit simultaneous laser pulses at the different wavelengths. The reflected multispectral energy is captured through a receiver of the sensor, and the return signal together with the position and orientation information of sensor is recorded. These recorded data are solved with GNSS/IMU data for further post-processing, forming high density multispectral 3D point clouds. As the first commercial multispectral airborne Lidar sensor, Optech Titan system is capable of collecting point clouds data from all three channels at 532nm visible (Green), at 1064 nm near infrared (NIR) and at 1550nm intermediate infrared (IR). It has become a new source of data for 3D land cover classification. The paper presents an Object Based Image Analysis (OBIA) approach to only use multispectral Lidar point clouds datasets for 3D land cover classification. The approach consists of three steps. Firstly, multispectral intensity images are segmented into image objects on the basis of multi-resolution segmentation integrating different scale parameters. Secondly, intensity objects are classified into nine categories by using the customized features of classification indexes and a combination the multispectral reflectance with the vertical distribution of object features. Finally, accuracy assessment is conducted via comparing random reference samples points from google imagery tiles with the classification results. The classification results show higher overall accuracy for most of the land cover types. Over 90% of overall accuracy is achieved via using multispectral Lidar point clouds for 3D land cover classification.
Purpose:To explore the feasibility and repeatability of a novel glasses-free display combined with random-dot stimulus and eye-tracking technology for screening stereopsis in adults.Methods:A total of 74 patients aged 18–44 years were recruited in this study (male: female, 32:42), including 33 patients with high myopia [≤ -6.0 diopters (D)] and 41 patients with moderate-to-low myopia (>-6.0 D). Stereopsis was measured using glasses-free, polarized, and Titmus stereotests. All patients completed a visual fatigue questionnaire after the polarized stereotest and glasses-free test. Kendall's W and Cohen's Kappa tests were used to evaluate repeatability and consistency of the glasses-free stereotest.Results:The stereotest results using the glasses-free monitor showed strong repeatability in the three consecutive tests (W = 0.968, P < 0.01) and good consistency with the polarized stereotest and Titmus test results (vs. polarization: Kappa = 0.910, P < 0.001; vs. Titmus: Kappa = 0.493, P < 0.001). Stereopsis levels of the high myopia group were significantly poorer than those of the moderate-to-low myopia group in three stereotest monitors (all P < 0.05). There was no significant difference in visual fatigue level between the polarized and the glasses-free display test (P = 0.72). Compared with the polarized test, 56.76% of patients preferred the glasses-free display and found it more comfortable, 20.27% reported both tests to be acceptable.Conclusions:In our adult patients, the new eye-tracking glasses-free display system feasibly screened stereopsis with good repeatability, consistency, and patient acceptance.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.