Head pose estimation using image abstraction and local directional quaternary patterns for multiclass classification

Han, ByungOk; Lee, Suwon; Yang, Hyun Seung

doi:10.1016/j.patrec.2014.03.017

Cited by 21 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another approach reported in [22] uses reflection symmetry information in covariant features extracted from Gabor features. Features derived from local directional quaternary patterns (LDQP) have been used in conjunction with linear SVM successfully in high resolution RGB data [23].…”

Section: A Head Pose Estimationmentioning

confidence: 99%

See 1 more Smart Citation

Deep Head Pose: Gaze-Direction Estimation in Multimodal Video

Mukherjee

Robertson

2015

IEEE Trans. Multimedia

173

View full text Add to dashboard Cite

In this paper we present a Convolutional Neural Network based model for human head pose estimation in low-resolution multi-modal RGB-D data. We pose the problem as one of classification of human gazing direction. We further fine-tune a regressor based on the learned deep classifier. Next we combine the two models (classification and regression) to estimate approximate regression confidence. We present state-ofthe-art results in datasets that span the range of high resolution Human Robot Interaction (close up faces plus depth information) data to challenging low resolution outdoor surveillance data. We build upon our robust head-pose estimation and further introduce a new visual attention model to recover interaction with the environment. Using this probabilistic model we show that many higher level scene understanding like human-human/scene interaction detection can be achieved. Our solution runs in real-time on commercial hardware.

show abstract

Section: A Head Pose Estimationmentioning

confidence: 99%

“…We compare our method against two state-of-the-art techniques on this dataset 1. LDQP [23] and 2. circle23Sphere [21]. As shown in Table I we outperform both competing techniques [23] and [21] in terms of Mean Angular Error(MAE) by a significant margin without any training on this dataset.…”

Section: ) Validation Onmentioning

confidence: 99%

Deep Head Pose: Gaze-Direction Estimation in Multimodal Video

Mukherjee

Robertson

2015

IEEE Trans. Multimedia

173

View full text Add to dashboard Cite

show abstract

“…Han et al [2] proposed the Image Abstraction and Local Directional Quaternary Pattern (IA-LDQP). An edgelike image is extracted from a precisely segmented-image using a Difference of Gaussian (DoG) filter.…”

Section: Related Workmentioning

confidence: 99%

“…Dataset DS1, DS2, DS3, DS4 and DS5 are used independently for offline evaluations. In this section, we compare our method with COG [4], IA-LDQP [2], WLD [5], and CovGA [3]. We re-implement COG, IA-LDQP, and WLD, while we used the result of CovGa directly from their paper.…”

Section: Datasetmentioning

confidence: 99%

Estimating Head Orientation Using a Combination of Multiple Cues

Dewantara

Miura

2016

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYThis paper proposes an appearance-based novel descriptor for estimating head orientation. Our descriptor is inspired by the Weberbased feature, which has been successfully implemented for robust texture analysis, and the gradient which performs well for shape analysis. To further enhance the orientation differences, we combine them with an analysis of the intensity deviation. The position of a pixel and its intrinsic intensity are also considered. All features are then composed as a feature vector of a pixel. The information carried by each pixel is combined using a covariance matrix to alleviate the influence caused by rotations and illumination. As the result, our descriptor is compact and works at high speed. We also apply a weighting scheme, called Block Importance Feature using Genetic Algorithm (BIF-GA), to improve the performance of our descriptor by selecting and accentuating the important blocks. Experiments on three head pose databases demonstrate that the proposed method outperforms the current state-of-the-art methods. Also, we can extend the proposed method by combining it with a head detection and tracking system to enable it to estimate human head orientation in real applications.

show abstract

“…In particular, people innately recognize the shapes, configurations, or contours of trained features such as eyes, noses, mouths, eyebrows, foreheads, and chins. Thus, people can remember abstracted images of heads by inference from trained data [7].…”

Section: Geometric Approachesmentioning

confidence: 99%

Head Pose Estimation Based On Detecting Facial Features

Hatem

Zou

Majeed

et al. 2015

IJMUE

View full text Add to dashboard Cite

Head pose estimation is recently a more popular area of research. Challenging conditions, such as extreme pose, lighting, and occlusion, has historically hampered traditional, model-based methods. This paper presents a proposal of an integrated method for head pose estimation based on face detection and tracking. This method first locates certain facial features and based on their relative locations determine the head pose, the head pose estimated using coordinates of both eyes and a mouth relative to the nose as the origin of the coordinate system. The nose position is set up as the origin. The coordinates of the other parts defined from the origin, the distance between the face parts normalized so that the coordinates are independent of the image size. For facial feature detection from the detected face region, Haar-like feature utilized along with AdaBoost learning, the Adaboost learning algorithm used for creating optimized learning data. From the experiments, the proposed approach shows robustness in face and facial feature detection and eventually produces better results in estimating head pose rather than simply using Haar-like feature for both face and facial feature detection. The computational cost is low because it uses only those three points.

show abstract

Head pose estimation using image abstraction and local directional quaternary patterns for multiclass classification

Cited by 21 publications

References 14 publications

Deep Head Pose: Gaze-Direction Estimation in Multimodal Video

Deep Head Pose: Gaze-Direction Estimation in Multimodal Video

Estimating Head Orientation Using a Combination of Multiple Cues

Head Pose Estimation Based On Detecting Facial Features

Contact Info

Product

Resources

About