Heedless Backbones

FLatten-PVT 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

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
FLatten-PVT-T	12.2	IN-1k : Sup. : 300	— : — : —	2.0	77.8
FLatten-PVT-S	21.7	IN-1k : Sup. : 300	— : — : —	4.0	81.7
FLatten-PVT-M	37.2	IN-1k : Sup. : 300	— : — : —	7.0	83.0
FLatten-PVT-L	50.6	IN-1k : Sup. : 300	— : — : —	10.4	83.4

model	params (m)	pretrain	finetune	gflops	IN-1k
FLatten-PVT-T	12.2	IN-1k : Sup. : 300	— : — : —	2.0	77.8/93.91
FLatten-PVT-S	21.7	IN-1k : Sup. : 300	— : — : —	4.0	81.7/—
FLatten-PVT-M	37.2	IN-1k : Sup. : 300	— : — : —	7.0	83.0/—
FLatten-PVT-L	50.6	IN-1k : Sup. : 300	— : — : —	10.4	83.4/—

COCO (val)

model	pretrain	head	train	gflops	mAP^b	AP^b₅₀	AP^b₇₅	mAP^b_s	mAP^b_m	mAP^b_l
FLatten-PVT-T	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	244.0	38.2	61.6	41.9	24.1	40.7	51.0

COCO (val)

model	pretrain	head	train	gflops	mAP^m	AP^m₅₀	AP^m₇₅	mAP^m_s	mAP^m_m	mAP^m_l
FLatten-PVT-T	IN-1k : Sup. : 300	Mask R-CNN	COCO (train) : 12	244.0	37.0	57.6	39.0	19.4	39.0	52.1

ADE20K (val)

model	pretrain	head	train	gflops	mIoU^ms	pAcc^ms	mAcc^ms	mIoU^ss	pAcc^ss	mAcc^ss
FLatten-PVT-T	IN-1k : Sup. : 300	Panoptic FPN	ADE20K (train) : 64 : 512	169.0	—	—	—	37.21	—	48.95