TACO: Training-free Sound Prompted Segmentation via Semantically Constrained Audio-visual CO-factorization
This addresses the problem of segmenting image regions based on audio prompts for applications in multimedia analysis, but it is incremental as it builds on existing pre-trained models and NMF techniques.
The paper tackles sound-prompted segmentation by proposing a training-free method using Non-negative Matrix Factorization to co-factorize audio and visual features, achieving state-of-the-art performance in unsupervised segmentation and significantly surpassing previous methods.
Large-scale pre-trained audio and image models demonstrate an unprecedented degree of generalization, making them suitable for a wide range of applications. Here, we tackle the specific task of sound-prompted segmentation, aiming to segment image regions corresponding to objects heard in an audio signal. Most existing approaches tackle this problem by fine-tuning pre-trained models or by training additional modules specifically for the task. We adopt a different strategy: we introduce a training-free approach that leverages Non-negative Matrix Factorization (NMF) to co-factorize audio and visual features from pre-trained models so as to reveal shared interpretable concepts. These concepts are passed on to an open-vocabulary segmentation model for precise segmentation maps. By using frozen pre-trained models, our method achieves high generalization and establishes state-of-the-art performance in unsupervised sound-prompted segmentation, significantly surpassing previous unsupervised methods.