Asahi Kasei Kabushiki Kaisha

Asahi Kasei Corporation

Keiji Shimada

Yoshihiro Noguchi

Ken Ishihara

Hideaki Sasahara

An object is to provide a face pose estimating device, a face pose estimating method, and a face pose estimating program that allow high accuracy estimation of a face pose even if a face organ cannot be detected from an image. A face organ detectorof the face pose estimating devicedetects a face organ from an image frame. A face pose candidate set generatorgenerates a face pose candidate set. A first similarity estimatorcomputes 3D-model-based similarities. A second similarity estimatorcomputes appearance-based similarities. A first likelihood estimatorcomputes a first likelihood corresponding to the 3D-model-based similarity. A second likelihood estimatorcomputes a second likelihood corresponding to the appearance-based similarities. An integrated likelihood estimatorcomputes an integrated likelihood by using the first and second likelihoods. A face pose estimatorestimates the face pose by using the integrated likelihood.

Face pose estimation device, face pose estimation method and face pose estimation program

A face portion detection device, a behavior content classification device, a speech content classification device, a car navigation system, a face direction classification device, a face portion detection device control program, a behavior content classification device control program, a face portion detection device control method, and a behavior content classification device control method are provided for appropriately classifying a behavior content of the object person from a captured image including the face of the object person. A speech section detection deviceincludes an image capturing unit, a data storage unit, an image processing unit, a lip region detection unit, feature extraction unit, and a speech section detection unit. The lip region detection unituses a dedicated SVM to detect a lip region from a captured image, and the speech section detection unituses features of an image of a detected lip region and a dedicated HMM to detect a speech section.

Growing community of inventors

Tokyo, Japan

Keiji Shimada