One-Shot Instance Segmentation

We tackle the problem of one-shot instance segmentation: Given an example image of a novel, previously unknown object category, find and segment all objects of this category within a complex scene. To address this challenging new task, we propose Siamese Mask R-CNN... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
One-Shot Instance Segmentation COCO Siamese Mask R-CNN AP 0.5 14.5 # 1
One-Shot Object Detection COCO Siamese Mask R-CNN AP 0.5 16.3 # 2

Methods used in the Paper


METHOD TYPE
Softmax
Output Functions
Convolution
Convolutions
RoIAlign
RoI Feature Extractors
Mask R-CNN
Instance Segmentation Models