WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose (BMVC2020) @KarolMajek

WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose (BMVC2020) @KarolMajek

Uploaded April 2021 | Updated September 2026, 3 weeks ago

Code:
github.com/Ascend-Research/HeadPoseEstimation-WHENet

Paper:
arxiv.org/abs/2005.10353

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

WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose (BMVC2020)

TensorRT YOLOv4 608x608 - RTX2070 laptop

GDarknet - YoloV3 with GIoU loss

Zero-Shot Object Detection

ATSS R50 FPN 1x Adaptive Training Sample Selection

Tensorflow Monodepth - City2eigen dataset - Tesla K40c

DETR ResNet 50 Object Detection

Tensorflow Monodepth model trained on Cityscapes

MegaDepth

OpenVINO SSDLite Mobilenet V2

YOLOv3 + Deep Sort tracking by yehengchen

Street signs, manholes, vents, air conditioners detection

WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose (BMVC2020)