MogaNet Family

Paper | GitHub

axis

Select an option

Results

Parameters (M)

Images / Second

Publication Date

task

Select an option

---------

Object Detection

Instance Segmentation

Classification

Semantic Segmentation

Panoptic Segmentation

dataset

Select an option

---------

Cityscapes (val)

Cityscapes (test)

ADE20K (val)

ADE20K (test)

metric

Select an option

mIoU^ms

pAcc^ms

mAcc^ms

mIoU^ss

pAcc^ss

mAcc^ss

GFLOPs

head

Select an option

---------

UPerNet

Mask2Former

Panoptic FPN

resolution

Select an option

----------

512x2048

640x2560

axis

Select an option

Results

Parameters (M)

Images / Second

GFLOPs

Publication Date

Filters

pretrain dataset

Select an option

---------

JFT-3B

ImageNet-1k

ImageNet-22k

JFT-300M

pretrain method

Select an option

----------

Supervised

FCMAE

MAE

Legend

Legend attribute

Select an option

Family

Pretrain Dataset

Pretrain Method

Semantic Segmentation Head

Semantic Segmentation Resolution

Semantic Segmentation Training Epochs

Legend attribute (second)

Select an option

----------

Family

Pretrain Method

Semantic Segmentation Head

Semantic Segmentation Resolution

Semantic Segmentation Training Epochs

No Results

model	params (m)	pretrain	finetune	gflops	IN-1k
MogaNet-XT	3.0	IN-1k : Sup. : 300	— : — : —	0.8	76.5/—
MogaNet-XT	3.0	IN-1k : Sup. : 300	IN-1k : 30 : 256	1.04	77.2/—
MogaNet-T	5.2	IN-1k : Sup. : 300	— : — : —	1.1	79.0/—
MogaNet-T	5.2	IN-1k : Sup. : 300	IN-1k : 30 : 256	1.44	80.0/—
MogaNet-S	25.3	IN-1k : Sup. : 300	— : — : —	4.97	83.4/—
MogaNet-B	43.8	IN-1k : Sup. : 300	— : — : —	9.93	84.3/—
MogaNet-L	82.5	IN-1k : Sup. : 300	— : — : —	15.9	84.7/—
MogaNet-XL	180.8	IN-1k : Sup. : 300	— : — : —	34.5	85.1/—
MogaNet-XL	180.8	IN-22k : Sup. : 90	IN-1k : 30 : 384	102.0	87.8/—

COCO (val)

model	pretrain	head	train	gflops	mAP^b	AP^b₅₀	AP^b₇₅	mAP^b_s	mAP^b_m	mAP^b_l
MogaNet-XT	IN-1k : Sup. : 300	RetinaNet	COCO (train) : 12	167.0	39.7	60.0	42.4	23.8	43.6	51.7
MogaNet-XT	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	185.0	40.7	62.3	44.4	—	—	—
MogaNet-T	IN-1k : Sup. : 300	RetinaNet	COCO (train) : 12	173.0	41.4	61.5	44.4	25.1	45.7	53.6
MogaNet-T	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	192.0	42.6	64.0	46.4	—	—	—
MogaNet-S	IN-1k : Sup. : 300	RetinaNet	COCO (train) : 12	253.0	45.8	66.6	49.0	29.1	50.1	59.8
MogaNet-S	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	272.0	46.7	68.0	51.3	—	—	—
MogaNet-S	IN-1k : Sup. : 300	Cascade Mask R-CNN	COCO (train) : 36	750.0	51.6	70.8	56.3	—	—	—
MogaNet-B	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	373.0	47.9	70.0	52.7	—	—	—
MogaNet-B	IN-1k : Sup. : 300	Cascade Mask R-CNN	COCO (train) : 36	851.0	52.6	72.0	57.3	—	—	—
MogaNet-L	IN-1k : Sup. : 300	Cascade Mask R-CNN	COCO (train) : 36	974.0	53.3	71.8	57.8	—	—	—
MogaNet-XL	IN-22k : Sup. : 90	Cascade Mask R-CNN	COCO (train) : 36	1355.0	56.2	75.0	61.2	—	—	—

COCO (val)

model	pretrain	head	train	gflops	mAP^m	AP^m₅₀	AP^m₇₅	mAP^m_s	mAP^m_m	mAP^m_l
MogaNet-XT	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	185.0	37.6	59.6	40.2	—	—	—
MogaNet-T	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	192.0	39.1	61.3	42.0	—	—	—
MogaNet-S	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	272.0	42.2	65.4	45.5	—	—	—
MogaNet-S	IN-1k : Sup. : 300	Cascade Mask R-CNN	COCO (train) : 36	750.0	45.1	68.7	48.8	—	—	—
MogaNet-B	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	373.0	43.2	67.0	46.6	—	—	—
MogaNet-B	IN-1k : Sup. : 300	Cascade Mask R-CNN	COCO (train) : 36	851.0	46.0	69.6	49.7	—	—	—
MogaNet-L	IN-1k : Sup. : 300	Cascade Mask R-CNN	COCO (train) : 36	974.0	46.1	69.2	49.8	—	—	—
MogaNet-XL	IN-22k : Sup. : 90	Cascade Mask R-CNN	COCO (train) : 36	1355.0	48.8	72.6	53.3	—	—	—

ADE20K (val)

model	pretrain	head	train	gflops	mIoU^ms	pAcc^ms	mAcc^ms	mIoU^ss	pAcc^ss	mAcc^ss
MogaNet-XT	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 128 : 512	856.0	—	—	—	42.2	—	—
MogaNet-T	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 128 : 512	862.0	—	—	—	43.7	—	—
MogaNet-S	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 128 : 512	946.0	—	—	—	49.2	—	—
MogaNet-B	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 128 : 512	1050.0	—	—	—	50.1	—	—
MogaNet-L	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 128 : 512	1176.0	—	—	—	50.9	—	—
MogaNet-XL	IN-22k : Sup. : 90	UPerNet	ADE20K (train) : 128 : 640	2451.0	—	—	—	54.0	—	—