BLSP-Emo: Towards Empathetic Large Speech-Language Models
This work addresses the problem of creating accessible empathetic AI for human-computer interaction, though it is incremental by building on existing methods for speech-language models.
The paper tackles the challenge of developing an empathetic end-to-end speech-language model by introducing BLSP-Emo, which uses a two-stage process with existing ASR and SER datasets to align semantics and emotions, resulting in a model that excels in understanding speech and generating empathetic responses in tasks and conversations.
The recent release of GPT-4o showcased the potential of end-to-end multimodal models, not just in terms of low latency but also in their ability to understand and generate expressive speech with rich emotions. While the details are unknown to the open research community, it likely involves significant amounts of curated data and compute, neither of which is readily accessible. In this paper, we present BLSP-Emo (Bootstrapped Language-Speech Pretraining with Emotion support), a novel approach to developing an end-to-end speech-language model capable of understanding both semantics and emotions in speech and generate empathetic responses. BLSP-Emo utilizes existing speech recognition (ASR) and speech emotion recognition (SER) datasets through a two-stage process. The first stage focuses on semantic alignment, following recent work on pretraining speech-language models using ASR data. The second stage performs emotion alignment with the pretrained speech-language model on an emotion-aware continuation task constructed from SER data. Our experiments demonstrate that the BLSP-Emo model excels in comprehending speech and delivering empathetic responses, both in instruction-following tasks and conversations.