YOLOv4 1024 FP16 - RTX2070 MaxQ - using ext_output @KarolMajek

YOLOv4 1024 FP16 - RTX2070 MaxQ - using ext_output @KarolMajek

Uploaded April 2020 | Updated September 2026, 2 weeks ago

Download my 4k video test sequence: archive.org/details/0002201705192

./darknet.exe detector demo ../../cfg/coco.data ../../cfg/yolov4-1024.cfg ../../yolov4.weights ~/Desktop/0002-20170519-2.mp4 -thresh 0.2 -ext_output -out_filename ~/Desktop/output.avi

YOLOv4 1024 FP16 - RTX2070 MaxQ - using ext_output

MMdetection: Cascade RCNN X101 64x4d FPN2x

Faster RCNN X101 64x4d FPN 1x

Faster RCNN HrNetV2p W32 1x

EfficientDet D1 VOC (toandaominh1997/EfficientDet.Pytorch)

MMdetection: Cascade Mask RCNN X101 32x4d FPN 1x

OpenCV YOLO Object Detection at Night - Adrian Rosebrock (pyImageSearch) implementation

F1/10 ESWEEK 2018 - 3min trials

YOLObile best14x 512x320

Cascade RCNN R101 FPN 1x (MMdetection)

Detectron2: RetinaNet R50 FPN 1x

Detectron2: Mask RCNN R50 FPN 3x gn - Instance segmentation

YOLOv4 1024 FP16 - RTX2070 MaxQ - using ext_output