Investigating and Enhancing Vision-Audio Capability in Omnimodal Large Language Models
This addresses a specific bottleneck in multimodal AI for applications requiring audio-visual understanding, but it is incremental as it builds on existing OLLM frameworks.
The paper tackled the problem of poor vision-audio integration in Omnimodal Large Language Models (OLLMs), where audio queries underperform compared to text queries, and proposed a Self-Knowledge Distillation method that improved performance on multimodal tasks by aligning vision-audio processing with vision-text capabilities.
Omnimodal Large Language Models (OLLMs) have shown significant progress in integrating vision and text, but still struggle with integrating vision and audio, often exhibiting suboptimal performance when processing audio queries compared to text queries. This disparity is primarily due to insufficient alignment between vision and audio modalities during training, leading to inadequate attention to visual information when using audio queries. To mitigate this issue, we propose a Self-Knowledge Distillation (Self-KD) training method where the vision-text component of the OLLM serves as the teacher and the vision-audio component as the student. This enables the model to process audio in a manner analogous to its text processing. Our experimental results demonstrate that Self-KD is an effective method for enhancing the vision-audio capabilities of OLLMs by learning from the vision-text components, which subsequently improves the interaction between audio and images and results in improved performance on multimodal tasks.