Heedless Backbones

VMINet 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

---------

ImageNet-1k

ImageNet-A

ImageNet-R

ImageNet-Sketch

ImageNet-C

ImageNet-C-bar

ImageNet-V2

ImageNet-ReaL

PASCAL VOC 2007 (val)

PASCAL VOC 2007 (test)

metric

Select an option

----------

Top-1

Top-5

GFLOPs

resolution

Select an option

----------

224x224

384x384

512x512

axis

Select an option

Results

Parameters (M)

Images / Second

GFLOPs

Publication Date

Filters

pretrain dataset

Select an option

---------

ImageNet-1k

ImageNet-22k

JFT-300M

JFT-3B

MegData73M

pretrain method

Select an option

----------

Supervised

Sup. + TL

FCMAE

MAE

CL

MAP

Legend

Legend attribute

Select an option

Family

Pretrain Dataset

Classification Resolution

Legend attribute (second)

Select an option

----------

Family

Pretrain Method

Classification Resolution

model	params (m)	pretrain	finetune	GFLOPs	Top-1
VMINet-Ti	2.0	IN-1k : Sup. : 300	— : — : —	0.3	70.7
VMINet-XS	7.0	IN-1k : Sup. : 300	— : — : —	1.4	78.6
VMINet-S	13.0	IN-1k : Sup. : 300	— : — : —	2.3	80.5
VMINet-B	28.0	IN-1k : Sup. : 300	— : — : —	4.8	82.4

model	params (m)	pretrain	finetune	gflops	IN-1k
VMINet-Ti	2.0	IN-1k : Sup. : 300	— : — : —	0.3	70.7/—
VMINet-XS	7.0	IN-1k : Sup. : 300	— : — : —	1.4	78.6/—
VMINet-S	13.0	IN-1k : Sup. : 300	— : — : —	2.3	80.5/—
VMINet-B	28.0	IN-1k : Sup. : 300	— : — : —	4.8	82.4/—

COCO (val)

model	pretrain	head	train	gflops	mAP^b	AP^b₅₀	AP^b₇₅	mAP^b_s	mAP^b_m	mAP^b_l
VMINet-XS	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	189.0	38.9	61.9	42.4	—	—	—
VMINet-S	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	201.0	43.2	65.3	47.3	—	—	—
VMINet-B	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	276.0	44.5	66.7	48.6	—	—	—

COCO (val)

model	pretrain	head	train	gflops	mAP^m	AP^m₅₀	AP^m₇₅	mAP^m_s	mAP^m_m	mAP^m_l
VMINet-XS	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	189.0	36.4	58.7	38.8	—	—	—
VMINet-S	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	201.0	39.3	62.2	42.3	—	—	—
VMINet-B	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	276.0	40.5	63.7	43.7	—	—	—

ADE20K (val)

model	pretrain	head	train	gflops	mIoU^ms	pAcc^ms	mAcc^ms	mIoU^ss	pAcc^ss	mAcc^ss
VMINet-XS	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 95 : 512	None	42.7	—	—	—	—	—
VMINet-S	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 95 : 512	None	44.8	—	—	—	—	—
VMINet-B	IN-1k : Sup. : 300	UPerNet	ADE20K (train) : 95 : 512	None	47.2	—	—	—	—	—