Hybrid Task Cascade, or HTC, is a framework for cascading in instance segmentation. It differs from Cascade Mask R-CNN in two important aspects: (1) instead of performing cascaded refinement on the two tasks of detection and segmentation separately, it interweaves them for a joint multi-stage processing; (2) it adopts a fully convolutional branch to provide spatial context, which can help distinguishing hard foreground from cluttered background.
Source: Hybrid Task Cascade for Instance SegmentationPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Text Classification | 10 | 16.95% |
Instance Segmentation | 7 | 11.86% |
Multi-Label Classification | 6 | 10.17% |
Semantic Segmentation | 6 | 10.17% |
Classification | 4 | 6.78% |
Object Detection | 4 | 6.78% |
Language Modelling | 2 | 3.39% |
Multi-Label Text Classification | 2 | 3.39% |
Document Embedding | 1 | 1.69% |
Component | Type |
|
---|---|---|
1x1 Convolution
|
Convolutions | |
Convolution
|
Convolutions | |
FPN
|
Feature Extractors | |
RoIAlign
|
RoI Feature Extractors | |
RPN
|
Region Proposal |