Layered Embeddings for Amodal Instance Segmentation
This work addresses the challenge of segmenting occluded objects in computer vision, which is important for applications like robotics and autonomous driving, and represents an incremental improvement over existing methods.
The paper tackles the problem of amodal instance segmentation by extending semantic instance segmentation to include both visible and occluded parts, using a fully convolutional network to produce layered embeddings for clustering, resulting in accurate complete mask estimation that outperforms top-down bounding-box approaches.
The proposed method extends upon the representational output of semantic instance segmentation by explicitly including both visible and occluded parts. A fully convolutional network is trained to produce consistent pixel-level embedding across two layers such that, when clustered, the results convey the full spatial extent and depth ordering of each instance. Results demonstrate that the network can accurately estimate complete masks in the presence of occlusion and outperform leading top-down bounding-box approaches. Source code available at https://github.com/yanfengliu/layered_embeddings