PanoSAMic: Panoramic Image Segmentation from SAM Feature Encoding and Dual View Fusion
This work addresses the problem of accurate segmentation in panoramic images for applications like virtual reality or robotics, representing an incremental improvement by adapting existing foundation models to a specialized domain.
The paper tackles semantic segmentation of panoramic images by integrating a modified SAM encoder with a novel spatio-modal fusion module and spherical attention, achieving state-of-the-art results on Stanford2D3DS and Matterport3D datasets.
Existing image foundation models are not optimized for spherical images having been trained primarily on perspective images. PanoSAMic integrates the pre-trained Segment Anything (SAM) encoder to make use of its extensive training and integrate it into a semantic segmentation model for panoramic images using multiple modalities. We modify the SAM encoder to output multi-stage features and introduce a novel spatio-modal fusion module that allows the model to select the relevant modalities and best features from each modality for different areas of the input. Furthermore, our semantic decoder uses spherical attention and dual view fusion to overcome the distortions and edge discontinuity often associated with panoramic images. PanoSAMic achieves state-of-the-art (SotA) results on Stanford2D3DS for RGB, RGB-D, and RGB-D-N modalities and on Matterport3D for RGB and RGB-D modalities. https://github.com/dfki-av/PanoSAMic