Self-Supervised Learning of Audio Representations using Angular Contrastive Loss | IEEE Conference Publication | IEEE Xplore