Perceiving the Invisible: Proposal-Free Amodal Panoptic Segmentation
This work addresses the problem of comprehensive scene understanding for autonomous driving systems by improving segmentation accuracy in occlusion scenarios, representing an incremental advance in domain-specific methods.
The paper tackles amodal panoptic segmentation by predicting semantic labels of visible regions and full shapes of traffic participants, including occluded parts, using a proposal-free framework and achieves new state-of-the-art results on BDD100K-APS and KITTI-360-APS datasets.
Amodal panoptic segmentation aims to connect the perception of the world to its cognitive understanding. It entails simultaneously predicting the semantic labels of visible scene regions and the entire shape of traffic participant instances, including regions that may be occluded. In this work, we formulate a proposal-free framework that tackles this task as a multi-label and multi-class problem by first assigning the amodal masks to different layers according to their relative occlusion order and then employing amodal instance regression on each layer independently while learning background semantics. We propose the \net architecture that incorporates a shared backbone and an asymmetrical dual-decoder consisting of several modules to facilitate within-scale and cross-scale feature aggregations, bilateral feature propagation between decoders, and integration of global instance-level and local pixel-level occlusion reasoning. Further, we propose the amodal mask refiner that resolves the ambiguity in complex occlusion scenarios by explicitly leveraging the embedding of unoccluded instance masks. Extensive evaluation on the BDD100K-APS and KITTI-360-APS datasets demonstrate that our approach set the new state-of-the-art on both benchmarks.