Interactive Segment Anything NeRF with Feature Imitation
This work addresses the problem of enabling real-time interaction with objects in complex 3D scenes for applications like VR and digital creation, representing an incremental improvement by integrating existing models.
The paper tackles the lack of semantics in Neural Radiance Fields (NeRF) for interactive applications by proposing a framework that imitates backbone features from perception models to achieve zero-shot semantic segmentation, accelerating segmentation by 16 times with comparable mask quality.
This paper investigates the potential of enhancing Neural Radiance Fields (NeRF) with semantics to expand their applications. Although NeRF has been proven useful in real-world applications like VR and digital creation, the lack of semantics hinders interaction with objects in complex scenes. We propose to imitate the backbone feature of off-the-shelf perception models to achieve zero-shot semantic segmentation with NeRF. Our framework reformulates the segmentation process by directly rendering semantic features and only applying the decoder from perception models. This eliminates the need for expensive backbones and benefits 3D consistency. Furthermore, we can project the learned semantics onto extracted mesh surfaces for real-time interaction. With the state-of-the-art Segment Anything Model (SAM), our framework accelerates segmentation by 16 times with comparable mask quality. The experimental results demonstrate the efficacy and computational advantages of our approach. Project page: \url{https://me.kiui.moe/san/}.