Update README.md

README.md

@@ -1,2 +1,11 @@
-# EmoSphere-TTS
-The official implementation of EmoSphere-TTS
+# EmoSphere-TTS: Emotional Style and Intensity Modeling via Spherical Emotion Vector for Controllable Emotional Text-to-Speech <br><sub>The official implementation of EmoSphere-TTS</sub>
+##  <a src="https://img.shields.io/badge/cs.CV-2406.07803-b31b1b?logo=arxiv&logoColor=red" href="https://arxiv.org/abs/2406.07803"> <img src="https://img.shields.io/badge/cs.CV-2406.07803-b31b1b?logo=arxiv&logoColor=red"></a>|[Demo page](https://emosphere-tts.github.io/)
+
+**Deok-Hyeon Cho, Hyung-Seok Oh, Seung-Bin Kim, Sang-Hoon Lee, Seong-Whan Lee**
+
+Department of Artificial Intelligence, Korea University, Seoul, Korea  
+
+## Abstract
+Despite rapid advances in the field of emotional text-to-speech (TTS), recent studies primarily focus on mimicking the average style of a particular emotion. As a result, the ability to manipulate speech emotion remains constrained to several predefined labels, compromising the ability to reflect the nuanced variations of emotion. In this paper, we propose EmoSphere-TTS, which synthesizes expressive emotional speech by using a spherical emotion vector to control the emotional style and intensity of the synthetic speech. Without any human annotation, we use the arousal, valence, and dominance pseudo-labels to model the complex nature of emotion via a Cartesian-spherical transformation. Furthermore, we propose a dual conditional adversarial network to improve the quality of generated speech by reflecting the multi-aspect characteristics. The experimental results demonstrate the model’s ability to control emotional style and intensity with high-quality expressive speech.
+
+![240312_model_overview_1](https://github.com/Choddeok/EmoSphere-TTS/assets/77186350/913610da-bfcc-4e60-b8fe-c1172b8dc154)