{"id":2376,"date":"2021-01-14T09:00:16","date_gmt":"2021-01-14T01:00:16","guid":{"rendered":"https:\/\/blog.ailia.ai\/%e6%9c%aa%e5%88%86%e9%a1%9e\/hope-net-machine-learning-model-for-head-pose-estimation\/"},"modified":"2025-05-14T15:31:48","modified_gmt":"2025-05-14T07:31:48","slug":"hope-net-machine-learning-model-for-head-pose-estimation","status":"publish","type":"post","link":"https:\/\/blog.ailia.ai\/en\/tips-en\/hope-net-machine-learning-model-for-head-pose-estimation\/","title":{"rendered":"HOPE-Net : A Machine Learning Model for Estimating Face Orientation"},"content":{"rendered":"\n

Overview<\/strong><\/p>\n\n\n\n

HOPE-Net <\/em>is a machine learning model released in October 2017 which compute the angles in three axes (yaw, pitch, and roll) of a face in an input image.<\/a><\/p>\n\n\n\n

Fine-Grained Head Pose Estimation Without Keypoints<\/a><\/p>\n\n\n\n

Architecture<\/strong><\/h3>\n\n\n\n

Face orientation detection is an important technology used in gaze detection and recognition of which objects is being watched in a scene.<\/p>\n\n\n\n

Face orientation detection usually works by detecting key points of the target face and converting those points from 2D to 3D using a standard head model. However, there is a problem that the result depends on the accuracy of the face key points, and the need for ad-hoc fitting.<\/p>\n\n\n\n

HOPE-Net<\/em> uses multi-loss convolutional neural networks<\/em> to detect the orientation of faces in a single shot. Using the face detected by the face detector as input, ResNet50 <\/em>extracts features and FC Layer calculates yaw, pitch, and roll.<\/p>\n\n\n\n

\"\"
\u51fa\u5178\uff1ahttps:\/\/arxiv.org\/pdf\/1710.00925<\/a><\/figcaption><\/figure>\n\n\n\n

HOPE-Net performs best on AFLW2000, a dataset made of the first 2000 images of the Annotated Facial Landmarks in the Wild (AFLW) <\/em>dataset, which have been re-annotated with 68 3D landmarks.<\/p>\n\n\n\n

\"\"\/
\u51fa\u5178\uff1ahttps:\/\/arxiv.org\/pdf\/1710.00925<\/a><\/figcaption><\/figure>\n\n\n\n

HOPE-Net Usage<\/strong><\/h3>\n\n\n\n

Use the following command to run HOPE-Net and detect face orientation from a web camera.<\/p>\n\n\n\n

$ python3 hopenet.py -v 0<\/code><\/pre>\n\n\n\n
You can also use a faster version that uses ShuffleNetV2 <\/em>instead of ResNet50 <\/em>with the following command.<\/p>\n\n\n\n
$ python3 blazehand.py --lite -v 0\n<\/a><\/code><\/pre>\n\n\n\n
axinc-ai\/ailia-models<\/a><\/p>\n\n\n\n
Here is the kind of result you can expect.<\/p>\n\n\n\n
\n