Heedless Backbones

CloFormer 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
CloFormer-XXS	4.2	IN-1k : Sup. : 300	— : — : —	0.6	77.0
CloFormer-XS	7.2	IN-1k : Sup. : 300	— : — : —	1.1	79.8
CloFormer-S	12.3	IN-1k : Sup. : 300	— : — : —	2.0	81.6

model	params (m)	pretrain	finetune	gflops	IN-1k
CloFormer-XXS	4.2	IN-1k : Sup. : 300	— : — : —	0.6	77.0/—
CloFormer-XS	7.2	IN-1k : Sup. : 300	— : — : —	1.1	79.8/—
CloFormer-S	12.3	IN-1k : Sup. : 300	— : — : —	2.0	81.6/—

COCO (val)

model	pretrain	head	train	gflops	mAP^b	AP^b₅₀	AP^b₇₅	mAP^b_s	mAP^b_m	mAP^b_l
CloFormer-XXS	IN-1k : Sup. : 300	RetinaNet	COCO (train) : 12	None	39.6	60.3	42.1	24.2	43.1	53.2
CloFormer-XXS	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	None	40.3	62.6	43.4	—	—	—
CloFormer-XS	IN-1k : Sup. : 300	RetinaNet	COCO (train) : 12	None	42.2	63.2	44.9	26.6	45.7	56.4
CloFormer-XS	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	None	42.9	65.1	47.1	—	—	—
CloFormer-S	IN-1k : Sup. : 300	RetinaNet	COCO (train) : 12	None	43.9	65.1	47.2	27.2	47.6	58.2
CloFormer-S	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	None	45.1	67.8	49.1	—	—	—

COCO (val)

model	pretrain	head	train	gflops	mAP^m	AP^m₅₀	AP^m₇₅	mAP^m_s	mAP^m_m	mAP^m_l
CloFormer-XXS	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	None	37.3	59.6	39.5	—	—	—
CloFormer-XS	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	None	39.2	61.6	42.4	—	—	—
CloFormer-S	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	None	41.4	64.7	44.2	—	—	—

ADE20K (val)

model	pretrain	head	train	gflops	mIoU^ms	pAcc^ms	mAcc^ms	mIoU^ss	pAcc^ss	mAcc^ss
CloFormer-XXS	IN-1k : Sup. : 300	Panoptic FPN	ADE20K (train) : 64 : 512	25.2	—	—	—	40.4	—	—
CloFormer-XS	IN-1k : Sup. : 300	Panoptic FPN	ADE20K (train) : 64 : 512	28.2	—	—	—	42.4	—	—
CloFormer-S	IN-1k : Sup. : 300	Panoptic FPN	ADE20K (train) : 64 : 512	34.6	—	—	—	44.2	—	—