OV-MER: Towards Open-Vocabulary Multimodal Emotion Recognition | Read Paper on Bytez