Fully Convolutional Instance-aware Semantic Segmentation

Yi Li, Haozhi Qi, Jifeng Dai, Xiangyang Ji, Yichen Wei
11/23/2016
cs.CV

Official code: msracver/FCIS (1,562 stars, MIT, last commit 2021-09-27).

Abstract

We present the first fully convolutional end-to-end solution for instance-aware semantic segmentation task. It inherits all the merits of FCNs for semantic segmentation and instance mask proposal. It performs instance mask prediction and classification jointly. The underlying convolutional representation is fully shared between the two sub-tasks, as well as between all regions of interest. The proposed network is highly integrated and achieves state-of-the-art performance in both accuracy and efficiency. It wins the COCO 2016 segmentation competition by a large margin. Code would be released at \url{https://github.com/daijifeng001/TA-FCN}.

View on arXivView PDF

Code Implementations(1)

msracver/FCISOfficial100%
MIT

Fully Convolutional Instance-aware Semantic Segmentation

1,562403C, C++Oct 6, 20165 years agoMIT

Similar Papers

Conditional Convolutions for Instance Segmentation

87% match1 repo3.5k2020

D-PLS: Decoupled Semantic Segmentation for 4D-Panoptic-LiDAR-Segmentation

87% match2025

Vocabulary-Free 3D Instance Segmentation with Vision and Language Assistant

87% match2024

ISBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolution

86% match1 repo1692023

Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers

85% match1 repo1.1k2020

PSGformer: Enhancing 3D Point Cloud Instance Segmentation via Precise Semantic Guidance

85% match2023

Cite this paper

@article{li2016fully,
  title  = {Fully Convolutional Instance-aware Semantic Segmentation},
  author = {Yi Li and Haozhi Qi and Jifeng Dai and Xiangyang Ji and Yichen Wei},
  year   = {2016},
  eprint = {1611.07709},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1611.07709}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1611.07709.svg)](https://codeofpaper.com/paper/1611.07709)

Discussion