Learning Category- and Instance-Aware Pixel Embedding for Fast Panoptic Segmentation
This work addresses the complex scene understanding problem of panoptic segmentation for computer vision researchers, offering a more efficient and streamlined approach.
The paper tackles panoptic segmentation by introducing a novel pixel embedding that encodes both category and instance information, enabling a simplified one-stage framework that achieves comparable performance to two-stage methods on COCO with fast inference speed.
Panoptic segmentation (PS) is a complex scene understanding task that requires providing high-quality segmentation for both thing objects and stuff regions. Previous methods handle these two classes with semantic and instance segmentation modules separately, following with heuristic fusion or additional modules to resolve the conflicts between the two outputs. This work simplifies this pipeline of PS by consistently modeling the two classes with a novel PS framework, which extends a detection model with an extra module to predict category- and instance-aware pixel embedding (CIAE). CIAE is a novel pixel-wise embedding feature that encodes both semantic-classification and instance-distinction information. At the inference process, PS results are simply derived by assigning each pixel to a detected instance or a stuff class according to the learned embedding. Our method not only demonstrates fast inference speed but also the first one-stage method to achieve comparable performance to two-stage methods on the challenging COCO benchmark.