knet

K-Net

K-Net: Towards Unified Image Segmentation

Introduction

Official Repo

Code Snippet

Abstract

Semantic, instance, and panoptic segmentations have been addressed using different and specialized frameworks despite their underlying connections. This paper presents a unified, simple, and effective framework for these essentially similar tasks. The framework, named K-Net, segments both instances and semantic categories consistently by a group of learnable kernels, where each kernel is responsible for generating a mask for either a potential instance or a stuff class. To remedy the difficulties of distinguishing various instances, we propose a kernel update strategy that enables each kernel dynamic and conditional on its meaningful group in the input image. K-Net can be trained in an end-to-end manner with bipartite matching, and its training and inference are naturally NMS-free and box-free. Without bells and whistles, K-Net surpasses all previous published state-of-the-art single-model results of panoptic segmentation on MS COCO test-dev split and semantic segmentation on ADE20K val split with 55.2% PQ and 54.3% mIoU, respectively. Its instance segmentation performance is also on par with Cascade Mask R-CNN on MS COCO with 60%-90% faster inference speeds. Code and models will be released at this https URL.

@inproceedings{zhang2021knet,
    title={{K-Net: Towards} Unified Image Segmentation},
    author={Wenwei Zhang and Jiangmiao Pang and Kai Chen and Chen Change Loy},
    year={2021},
    booktitle={NeurIPS},
}

Results and models

ADE20K

Method	Backbone	Crop Size	Lr schd	Mem (GB)	Inf time (fps)	mIoU	mIoU(ms+flip)	config	download
KNet + FCN	R-50-D8	512x512	80000	7.01	19.24	43.60	45.12	config	model | log
KNet + PSPNet	R-50-D8	512x512	80000	6.98	20.04	44.18	45.58	config	model | log
KNet + DeepLabV3	R-50-D8	512x512	80000	7.42	12.10	45.06	46.11	config	model | log
KNet + UPerNet	R-50-D8	512x512	80000	7.34	17.11	43.45	44.07	config	model | log
KNet + UPerNet	Swin-T	512x512	80000	7.57	15.56	45.84	46.27	config	model | log
KNet + UPerNet	Swin-L	512x512	80000	13.5	8.29	52.05	53.24	config	model | log
KNet + UPerNet	Swin-L	640x640	80000	18.31	5.55	52.46	53.78	config	model | log

Note:

• All experiments of K-Net are implemented with 8 V100 (32G) GPUs with 2 samplers per GPU.