Imagine to Hear: Auditory Knowledge Generation can be an Effective Assistant for Language Models
This addresses a specific limitation in language models for tasks requiring auditory knowledge, offering a more efficient alternative to database retrieval methods.
The paper tackles the problem of language models lacking auditory commonsense knowledge by proposing Imagine to Hear, a method that dynamically generates auditory knowledge using generative models instead of retrieving from external databases, achieving state-of-the-art performance on AuditoryBench.
Language models pretrained on text-only corpora often struggle with tasks that require auditory commonsense knowledge. Previous work addresses this problem by augmenting the language model to retrieve knowledge from external audio databases. This approach has several limitations, such as the potential lack of relevant audio in databases and the high costs associated with constructing the databases. To address these issues, we propose Imagine to Hear, a novel approach that dynamically generates auditory knowledge using generative models. Our framework detects multiple audio-related textual spans from the given prompt and generates corresponding auditory knowledge. We develop several mechanisms to efficiently process multiple auditory knowledge, including a CLAP-based rejection sampler and a language-audio fusion module. Our experiments show that our method achieves state-of-the-art performance on AuditoryBench without relying on external databases, highlighting the effectiveness of our generation-based approach.