Amodal Instance Segmentation
13 papers with code • 1 benchmarks • 2 datasets
Different from traditional segmentation which only focuses on visible regions, amodal instance segmentation also predicts the occluded parts of object instances.
Description Credit: Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers, CVPR'21
Libraries
Use these libraries to find Amodal Instance Segmentation models and implementations
WALT
MOViD-Amodal
Latest papers with no code
ShapeFormer: Shape Prior Visible-to-Amodal Transformer-based Amodal Instance Segmentation
Consequently, this compromised quality of visible features during the subsequent visible-to-amodal transition.
MUVA: A New Large-Scale Benchmark for Multi-View Amodal Instance Segmentation in the Shopping Scenario
To bridge this gap, we propose a new task called Multi-view Amodal Instance Segmentation (MAIS) and introduce the MUVA dataset, the first MUlti-View AIS dataset that takes the shopping scenario as instantiation.
Learning Vector Quantized Shape Code for Amodal Blastomere Instance Segmentation
To alleviate the problem, we propose to classify input features into intermediate shape codes and recover complete object shapes from them.
Amodal Instance Segmentation
We consider the problem of amodal instance segmentation, the objective of which is to predict the region encompassing both visible and occluded parts of each object.
