Detectron2: Faster RCNN R101 DC5 3x - COCO - object detection Tesla V100 @KarolMajek

Detectron2: Faster RCNN R101 DC5 3x - COCO - object detection Tesla V100 @KarolMajek

Uploaded January 2020 | Updated September 2026, 3 weeks ago

source code:
github.com/facebookresearch/detectron2

model:
dl.fbaipublicfiles.com/detectron2/COCO-Detection/faster_rcnn_R_101_DC5_3x/138204841/model_final_3e0943.pkl

Detectron2: Faster RCNN R101 DC5 3x - COCO - object detection Tesla V100

Tensorflow Monodepth - Eigen dataset - GeForce GTX 1080

YOLOv4 1024 - saving results to jpgs is slow

RefineDet512 ResNet101

YOLOv4 1024 RTX2070MaxQ (laptop) screencast

NanoDet-m 416 - RTX2070 laptop FP16

Darknet Nightmare

YOLOv3 608x608 CSP ResneXt50 panet spp original optimal (AlexeyAB)

Detectron2: Mask RCNN R50 FPN 1x

Cascade RCNN R50 FPN 1x

Tensorflow Monodepth - Cityscapes dataset - Tesla K40c

Tensorflow Monodepth - Cityscapes dataset - GeForce GTX TITAN X

Detectron2: Faster RCNN R101 DC5 3x - COCO - object detection Tesla V100