TASK	DATASET	MODEL	METRIC NAME	METRIC VALUE	GLOBAL RANK
Instance Segmentation	BDD100K	HRNet	AP	22.5	# 2
Instance Segmentation	BDD100K val	HRNet	AP	22.5	# 2
Semantic Segmentation	Cityscapes test	HRNetV2 (train+val)	Mean IoU (class)	81.6	# 30
Semantic Segmentation	Cityscapes val	HRNetV2 (HRNetV2-W48)	mIoU	81.1	# 19
Object Detection	COCO minival	HTC (HRNetV2p-W48)	box AP	47.0	# 57
Object Detection	COCO minival	HTC (HRNetV2p-W48)	APS	28.8	# 18
Object Detection	COCO minival	HTC (HRNetV2p-W48)	APM	50.3	# 16
Object Detection	COCO minival	HTC (HRNetV2p-W48)	APL	62.2	# 16
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W48, cascade)	box AP	46.0	# 65
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W48, cascade)	APS	27.5	# 24
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W48, cascade)	APL	60.1	# 20
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32, cascade)	APS	26.1	# 36
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32, cascade)	APM	47.9	# 27
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W18)	box AP	39.2	# 138
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W18)	APM	41.7	# 67
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W18)	APL	51.0	# 65
Instance Segmentation	COCO minival	HTC (HRNetV2p-W48)	mask AP	41.0	# 40
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W48)	box AP	41.8	# 109
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W48)	AP50	62.8	# 44
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W48)	AP75	45.9	# 49
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W48)	APM	44.7	# 48
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W48)	APL	54.6	# 51
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	box AP	44.6	# 78
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	AP50	62.7	# 46
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	AP75	48.7	# 28
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	APS	26.3	# 35
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	APM	48.1	# 24
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	APL	58.5	# 32
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	box AP	41.3	# 114
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	AP50	59.2	# 74
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	AP75	44.9	# 55
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	APS	23.7	# 55
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	APM	44.2	# 53
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	APL	54.1	# 52
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	box AP	40.9	# 118
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	AP50	61.8	# 52
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	AP75	44.8	# 56
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	APS	24.4	# 50
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	APM	43.7	# 57
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	APL	53.3	# 55
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	box AP	38.0	# 147
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	AP50	58.9	# 78
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	AP75	41.5	# 74
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	APS	22.6	# 63
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	APM	40.8	# 70
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	APL	49.6	# 70
Object Detection	COCO minival	HTC (HRNetV2p-W32)	box AP	45.3	# 69
Object Detection	COCO minival	HTC (HRNetV2p-W32)	APS	27.0	# 26
Object Detection	COCO minival	HTC (HRNetV2p-W32)	APM	48.4	# 20
Object Detection	COCO minival	HTC (HRNetV2p-W32)	APL	59.5	# 24
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	box AP	43.7	# 87
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	AP50	61.7	# 53
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	AP75	47.7	# 34
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	APS	25.6	# 41
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	APM	46.5	# 36
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	APL	57.4	# 39
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32)	box AP	42.3	# 103
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32)	APS	25.0	# 47
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32)	APM	45.4	# 46
Object Detection	COCO minival	HTC (HRNetV2p-W18)	box AP	43.1	# 95
Object Detection	COCO minival	HTC (HRNetV2p-W18)	APS	26.6	# 32
Object Detection	COCO minival	HTC (HRNetV2p-W18)	APM	46.0	# 39
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	box AP	40.5	# 157
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	AP50	59.3	# 134
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	APS	23.4	# 114
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	APM	42.6	# 132
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	APL	51.0	# 135
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	Hardware Burden	16G	# 1
Object Detection	COCO test-dev	FCOS (HRNetV2p-W48)	Operations per network pass	27.3G	# 1
Object Detection	COCO test-dev	Cascade R-CNN (HRNetV2p-W48)	AP75	48.6	# 80
Object Detection	COCO test-dev	Cascade R-CNN (HRNetV2p-W48)	APS	26.0	# 90
Object Detection	COCO test-dev	Cascade R-CNN (HRNetV2p-W48)	APM	47.3	# 87
Object Detection	COCO test-dev	Cascade R-CNN (HRNetV2p-W48)	APL	56.3	# 91
Object Detection	COCO test-dev	Cascade R-CNN (HRNetV2p-W48)	Hardware Burden	None	# 1
Object Detection	COCO test-dev	Cascade R-CNN (HRNetV2p-W48)	Operations per network pass	None	# 1
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	box AP	42.4	# 140
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	AP50	63.6	# 95
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	AP75	46.4	# 106
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	APS	24.9	# 98
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	APM	44.6	# 117
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	APL	53.0	# 122
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	Hardware Burden	16G	# 1
Object Detection	COCO test-dev	Faster R-CNN (HRNetV2p-W48)	Operations per network pass	20.8G	# 1
Object Detection	COCO test-dev	CenterNet (HRNetV2-W48)	box AP	43.5	# 123
Object Detection	COCO test-dev	CenterNet (HRNetV2-W48)	AP75	46.5	# 104
Object Detection	COCO test-dev	CenterNet (HRNetV2-W48)	APS	22.2	# 123
Object Detection	COCO test-dev	CenterNet (HRNetV2-W48)	APL	57.8	# 74
Object Detection	COCO test-dev	CenterNet (HRNetV2-W48)	Hardware Burden	16G	# 1
Object Detection	COCO test-dev	CenterNet (HRNetV2-W48)	Operations per network pass	21.7G	# 1
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W32 + cascade)	AP50	62.5	# 106
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W32 + cascade)	AP75	48.6	# 80
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W32 + cascade)	APL	56.3	# 91
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W32 + cascade)	Hardware Burden	16G	# 1
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W32 + cascade)	Operations per network pass	50.6G	# 1
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	box AP	47.3	# 84
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	AP50	65.9	# 69
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	AP75	51.2	# 61
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	APS	28.0	# 69
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	APM	49.7	# 66
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	APL	59.8	# 63
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	Hardware Burden	15G	# 1
Object Detection	COCO test-dev	HTC (HRNetV2p-W48)	Operations per network pass	71.7G	# 1
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	box AP	46.1	# 95
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	AP50	64.0	# 91
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	AP75	50.3	# 73
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	APS	27.1	# 77
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	APM	48.6	# 75
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	APL	58.3	# 69
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	Hardware Burden	15G	# 1
Object Detection	COCO test-dev	Mask R-CNN (HRNetV2p-W48 + cascade)	Operations per network pass	61.8G	# 1
Semantic Segmentation	DADA-seg	HRNet (ACDC)	mIoU	27.5	# 9
Thermal Image Segmentation	MFN Dataset	HRNet	mIOU	51.7	# 9
Semantic Segmentation	PASCAL Context	HRNetV2 HRNetV2-W48	mIoU	54	# 24
Semantic Segmentation	PASCAL Context	CFNet (ResNet-101)	mIoU	54.0	# 24

Badge	Markdown
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/object-detection-on-coco)](https://paperswithcode.com/sota/object-detection-on-coco?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/instance-segmentation-on-bdd100k)](https://paperswithcode.com/sota/instance-segmentation-on-bdd100k?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/instance-segmentation-on-bdd100k-val)](https://paperswithcode.com/sota/instance-segmentation-on-bdd100k-val?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/semantic-segmentation-on-dada-seg)](https://paperswithcode.com/sota/semantic-segmentation-on-dada-seg?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/thermal-image-segmentation-on-mfn-dataset)](https://paperswithcode.com/sota/thermal-image-segmentation-on-mfn-dataset?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/semantic-segmentation-on-cityscapes-val)](https://paperswithcode.com/sota/semantic-segmentation-on-cityscapes-val?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/semantic-segmentation-on-pascal-context)](https://paperswithcode.com/sota/semantic-segmentation-on-pascal-context?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/object-detection-on-coco-minival)](https://paperswithcode.com/sota/object-detection-on-coco-minival?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/semantic-segmentation-on-cityscapes)](https://paperswithcode.com/sota/semantic-segmentation-on-cityscapes?p=190807919)`
	`[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/190807919/instance-segmentation-on-coco-minival)](https://paperswithcode.com/sota/instance-segmentation-on-coco-minival?p=190807919)`

Deep High-Resolution Representation Learning for Visual Recognition

20 Aug 2019 · Jingdong Wang, Ke Sun, Tianheng Cheng, Borui Jiang, Chaorui Deng, Yang Zhao, Dong Liu, Yadong Mu, Mingkui Tan, Xinggang Wang, Wenyu Liu, Bin Xiao ·

High-resolution representations are essential for position-sensitive vision problems, such as human pose estimation, semantic segmentation, and object detection. Existing state-of-the-art frameworks first encode the input image as a low-resolution representation through a subnetwork that is formed by connecting high-to-low resolution convolutions \emph{in series} (e.g., ResNet, VGGNet), and then recover the high-resolution representation from the encoded low-resolution representation. Instead, our proposed network, named as High-Resolution Network (HRNet), maintains high-resolution representations through the whole process. There are two key characteristics: (i) Connect the high-to-low resolution convolution streams \emph{in parallel}; (ii) Repeatedly exchange the information across resolutions. The benefit is that the resulting representation is semantically richer and spatially more precise. We show the superiority of the proposed HRNet in a wide range of applications, including human pose estimation, semantic segmentation, and object detection, suggesting that the HRNet is a stronger backbone for computer vision problems. All the codes are available at~{\url{https://github.com/HRNet}}.

PDF Abstract

Code

Add Remove Mark official

open-mmlab/mmdetection

19,779

PaddlePaddle/PaddleDetection

7,725

PaddlePaddle/PaddleSeg

4,808

CSAILVision/semantic-segmentation-p…

↳ Quickstart in

Colab

4,318

open-mmlab/mmsegmentation

↳ Quickstart in

Colab

3,754

See all 31 implementations

Tasks

Add Remove

Instance Segmentation

Object Detection

Pose Estimation

Representation Learning

Semantic Segmentation

Thermal Image Segmentation

Datasets

COCO

Cityscapes

PASCAL Context

BDD100K

DADA-seg

Results from the Paper

Edit

Ranked #1 on Object Detection on COCO test-dev (Hardware Burden metric)

Get a GitHub badge

Task	Dataset	Model	Metric Name	Metric Value	Global Rank	Benchmark
Instance Segmentation	BDD100K	HRNet	AP	22.5	# 2	Compare
Instance Segmentation	BDD100K val	HRNet	AP	22.5	# 2	Compare
Semantic Segmentation	Cityscapes test	HRNetV2 (train+val)	Mean IoU (class)	81.6	# 30	Compare
Semantic Segmentation	Cityscapes val	HRNetV2 (HRNetV2-W48)	mIoU	81.1	# 19	Compare
Object Detection	COCO minival	HTC (HRNetV2p-W48)	box AP	47.0	# 57	Compare
			APS	28.8	# 18	Compare
			APM	50.3	# 16	Compare
			APL	62.2	# 16	Compare
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W48, cascade)	box AP	46.0	# 65	Compare
			APS	27.5	# 24	Compare
			APL	60.1	# 20	Compare
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32, cascade)	APS	26.1	# 36	Compare
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W32, cascade)	APM	47.9	# 27	Compare
Object Detection	COCO minival	Mask R-CNN (HRNetV2p-W18)	box AP	39.2	# 138	Compare
			APM	41.7	# 67	Compare
			APL	51.0	# 65	Compare
Instance Segmentation	COCO minival	HTC (HRNetV2p-W48)	mask AP	41.0	# 40	Compare
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W48)	box AP	41.8	# 109	Compare
			AP50	62.8	# 44	Compare
			AP75	45.9	# 49	Compare
			APM	44.7	# 48	Compare
			APL	54.6	# 51	Compare
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W48)	box AP	44.6	# 78	Compare
			AP50	62.7	# 46	Compare
			AP75	48.7	# 28	Compare
			APS	26.3	# 35	Compare
			APM	48.1	# 24	Compare
			APL	58.5	# 32	Compare
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W18)	box AP	41.3	# 114	Compare
			AP50	59.2	# 74	Compare
			AP75	44.9	# 55	Compare
			APS	23.7	# 55	Compare
			APM	44.2	# 53	Compare
			APL	54.1	# 52	Compare
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W32)	box AP	40.9	# 118	Compare
			AP50	61.8	# 52	Compare
			AP75	44.8	# 56	Compare
			APS	24.4	# 50	Compare
			APM	43.7	# 57	Compare
			APL	53.3	# 55	Compare
Object Detection	COCO minival	Faster R-CNN (HRNetV2p-W18)	box AP	38.0	# 147	Compare
			AP50	58.9	# 78	Compare
			AP75	41.5	# 74	Compare
			APS	22.6	# 63	Compare
			APM	40.8	# 70	Compare
			APL	49.6	# 70	Compare
Object Detection	COCO minival	HTC (HRNetV2p-W32)	box AP	45.3	# 69	Compare
			APS	27.0	# 26	Compare
			APM	48.4	# 20	Compare
			APL	59.5	# 24	Compare
Object Detection	COCO minival	Cascade R-CNN (HRNetV2p-W32)	box AP	43.7	# 87

Edit Social Preview

Deep High-Resolution Representation Learning for Visual Recognition

Code Edit Add Remove Mark official

Tasks Edit Add Remove

Datasets Edit

Results from the Paper Edit

Code

Add Remove Mark official

Tasks

Add Remove

Datasets

Results from the Paper

Edit