ATSS

Last updated on Feb 23, 2021

ATSS (R-101, 1x, pytorch)

Memory (M) 5600.0
inference time (s/im) 0.0813
File Size 196.16 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture ResNet, ATSS
lr sched 1x
Memory (M) 5600.0
Backbone Layers 101
inference time (s/im) 0.0813
SHOW MORE
SHOW LESS
ATSS (R-50, 1x, pytorch)

Memory (M) 3700.0
inference time (s/im) 0.05076
File Size 123.46 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture ResNet, ATSS
lr sched 1x
Memory (M) 3700.0
Backbone Layers 50
inference time (s/im) 0.05076
SHOW MORE
SHOW LESS
README.md

Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection

Introduction

[ALGORITHM]

@article{zhang2019bridging,
  title   =  {Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection},
  author  =  {Zhang, Shifeng and Chi, Cheng and Yao, Yongqiang and Lei, Zhen and Li, Stan Z.},
  journal =  {arXiv preprint arXiv:1912.02424},
  year    =  {2019}
}

Results and Models

Backbone Style Lr schd Mem (GB) Inf time (fps) box AP Config Download
R-50 pytorch 1x 3.7 19.7 39.4 config model | log
R-101 pytorch 1x 5.6 12.3 41.5 config model | log

Results

Object Detection on COCO minival

Object Detection
BENCHMARK MODEL METRIC NAME METRIC VALUE GLOBAL RANK
COCO minival ATSS (R-101, 1x, pytorch) box AP 41.5 # 60
COCO minival ATSS (R-50, 1x, pytorch) box AP 39.4 # 80