Heedless Backbones

Mamba-Adaptor 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

---------

COCO (val)

COCO (test)

Cityscapes (val)

Cityscapes (test)

ADE20K (val)

ADE20K (test)

metric

Select an option

----------

mAP

AP₅₀

AP₇₅

mAP_s

mAP_m

mAP_l

GFLOPs

head

Select an option

---------

Mask R-CNN

Cascade Mask R-CNN

Mask2Former

HTC++

HTC

Panoptic FPN

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

Instance Head

Instance Training Epochs

Legend attribute (second)

Select an option

----------

Family

Pretrain Method

Instance Head

Instance Training Epochs

model	params (m)	pretrain	head	train	GFLOPs	mAP
Mamba-Adaptor-b1	7.8	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	218.0	39.5
Mamba-Adaptor-b1	7.8	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 36	218.0	41.2
Mamba-Adaptor-b2	32.0	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	259.0	43.4
Mamba-Adaptor-b2	32.0	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 36	259.0	44.8

model	params (m)	pretrain	finetune	gflops	IN-1k
Mamba-Adaptor-b1	7.8	IN-1k : Sup. : 300	— : — : —	1.4	78.4/—
Mamba-Adaptor-b2	32.0	IN-1k : Sup. : 300	— : — : —	5.4	83.0/—

COCO (val)

model	pretrain	head	train	gflops	mAP^b	AP^b₅₀	AP^b₇₅	mAP^b_s	mAP^b_m	mAP^b_l
Mamba-Adaptor-b1	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	218.0	43.2	65.5	47.7	—	—	—
Mamba-Adaptor-b1	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 36	218.0	45.1	67.2	49.4	—	—	—
Mamba-Adaptor-b2	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	259.0	47.3	69.8	52.3	—	—	—
Mamba-Adaptor-b2	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 36	259.0	49.1	71.5	54.1	—	—	—

COCO (val)

model	pretrain	head	train	gflops	mAP^m	AP^m₅₀	AP^m₇₅	mAP^m_s	mAP^m_m	mAP^m_l
Mamba-Adaptor-b1	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	218.0	39.5	60.1	42.7	—	—	—
Mamba-Adaptor-b1	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 36	218.0	41.2	61.9	43.8	—	—	—
Mamba-Adaptor-b2	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	259.0	43.4	66.9	46.4	—	—	—
Mamba-Adaptor-b2	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 36	259.0	44.8	67.3	48.3	—	—	—