Fig 9 - uploaded by Naseem Ahmad
Content may be subject to copyright.
Source publication
Eye centre localization plays a crucial role in computer vision applications like face recognition, gaze estimation, driver fatigue detection, liveness detection, etc. However, it is difficult to localize the eye centre due to the variations in pose, occlusion, illumination, specular reflection, rotation, scale, etc. This work proposes an integrate...
Context in source publication
Context 1
... the face considered a region of interest. Faster RCNN will only look into the face to detect the eyes in this proposed method. Faster RCNN with ROI provides 97.32% and 97.49% accuracy for eye detection with $0.177 and $0.169 s for BioID and GI4E databases, respectively. The face and eye detection performance for both the databases is shown in Fig. 9 and Fig. 10. It has been observed that the BioID database is more challenging for face and eye ...
Similar publications
As a common method of deep learning, a convolutional neural network (CNN) shows excellent performance in face recognition. The features extracted by traditional face recognition methods are greatly influenced by subjective factors and are time-consuming and laborious. In addition, these images are susceptible to illumination, expression, occlusion,...
Citations
... Naseem et al. [18] adopt a faster RCNN deep learning model and AlexNet as the detection key for face, eyes and eye openness detection. The localization step combines techniques composed of rectangular-intensity-gradient approach. ...
Eye movements offer precious information about persons' state. Video surveillance, marketing, driver fatigue as well as medical diagnosis assistance applications manage eye behavior. We propose a new method for efficiently detecting eye movement. In this paper, we combine circle eye model with eye feature method to improve the accuracy. A set of detectors estimate the eyes centers to increase the localization rate. As a pre-processing stage, the mean of the edges yields the center of the two eye regions. Image treatment operations reduce the ROI. A Circle Hough Transform (CHT) algorithm is adopted in a modified version as a detector to find the circle eye in the image; the circle center found represents the eye's pupil estimation. We introduced the Maximally Stable Extremal Region (MSER) as a second detector, which has never been used for eye localization. Invariant to continuous geometric transformations and affine intensity changes and detected at several scales, MSERs efficiently detect regions of interest, in our case eye regions, and precisely, their centers. Ellipses fit MSERs, and their centroid estimation match eyes center. We demonstrate that the true eye centers can be found by combining these methods. The validation of the proposed method is performed on a very challenging BioID base. The proposed approach compares well with existing state-of-the-art techniques and achieves an accuracy of 82.53% on the BioID database when the normalized error is less than 0.05, without prior knowledge or any learning model.
... Naseem et al. (30) suggested Faster RCNN, and a rectangular-intensity-gradient technique for face, eye detection, openness identification, and eye center localization, with suggested AlexNet assisting in detecting eye state (opened or closed). On the other side, Ahmad and Laskar (2) used support vector machines (SVMs), and convolutional neural networks (CNNs)for iris-shape feature extraction to propose an accurate eye center identification and localization model. ...
The study of gaze tracking is a significant research area in computer vision. It focuses on real-world applications and the interface between humans and computers. Recently, new eye-tracking applications have boosted the need for low-cost methods. The eye region is a crucial aspect of tracking the direction of the gaze. In this paper, several new methods have been proposed for eye-tracking by using methods to determine the eye area as well as find the direction of gaze. Unmodified webcams can be used for eye-tracking without the need for specialized equipment or software. Two methods for determining the eye region were used: facial landmarks or the Haar cascade technique. Moreover, the direct method, based on the convolutional neural network model, and the engineering method, based on distances determining the iris region, were used to determine the eye’s direction. The paper uses two engineering techniques: drawing perpendicular lines on the iris region to identify the gaze direction junction point and dividing the eye region into five regions, with the blackest region representing the gaze direction. The proposed network model has proven effective in determining the eye’s gaze direction within limited mobility, while engineering methods improve their effectiveness in wide mobility.
... Utilizing these intrusive characteristics, most researchers localize the iris centre using handcrafted features. Therefore, circle fitting techniques such as circular HOG, circular geometric algebra (CGA) (Ma et al., 2020), ISF, rectangular-intensity-gradient (RIG) (Ahmad et al., 2022), and so forth, are used for iris detection. Gradient-based methods use dot products between gradient and displacement vectors for iris centre localization. ...
... Nsaif et al. (2021) proposed rough eye detection using Faster RCNN followed by a Gabor filter and a Naïve Bayes classifier for detecting true eyes. Ahmad et al. (2022) proposed a two-step approach for eye detection. Initially, they detect the face using a face detection algorithm. ...
... Hsu and Chung (2021) used an iris ripple filter with gradient estimation for iris centre localization. Ahmad et al. (2022) used a Faster RCNN for eye detection and a rectangular intensity-gradient method for iris centre localization. They used face detection before eye detection, and their method failed due to the glass frame, heavy occlusion, and hairs falling on the eyes. ...
Accurate iris centre localization is crucial in many computer vision and facial biometric applications such as gaze estimation, human–computer interaction, iris recognition, and liveness detection. However, it is challenging in an uncontrolled environment due to variations like pose, scale, rotation, specular reflection, and image quality. Therefore, a cascaded deep learning framework for iris centre localization in facial images is proposed that is robust to the abovementioned variations. The proposed approach consists of (i) YOLOv3 for eye detection, (ii) UNet for iris segmentation, and (iii) statistical modelling for iris centre localization. The eyes are first detected using the YOLOv3, and subsequently, iris segmentation is performed within the detected eyes using the UNet. Following iris segmentation, statistical modelling is employed to enhance the localization accuracy of the iris centre. Experiments were performed on benchmark databases, resulting in a standardized error measure SED of 3.405 pixels for BioID and 3.259 pixels for GI4E databases. In addition, the robustness of the proposed eye detection model was further evaluated on the Yale B for illumination variations and the CAS‐PEAL for pose variations.
... Their performance degrades in an uncontrolled environment. Researchers have recently been more focused on deep feature-based techniques in eye detection [9][10][11][12][13]. This is because deep features-based techniques can extract more information from facial images DCNNs and RCNNs are deep feature-based methods that utilize convolutional neural networks for image analysis and object detection. ...
... The Tiny-YOLOv3 model is a simplified model of YOLOv3 that maintain similar accuracy and faster processing time. Most researchers utilized face detectors [1,6,7,12] before detecting the eyes in a facial image. A face detector can help to reduce processing time and remove imposter-like images in the background. ...
... Iris localization is another intermediate step in iris tracking. In the context of iris detection, various techniques have been developed, including multi-scale ISF [6], sub-divided ISF [8], and rectangular-intensity-gradient (RIG) [12] features, etc. These techniques leverage the intensity and gradient information of the iris region. ...
Iris detection and tracking play an essential role in a wide range of real-world applications, including tracking gaze, biometric authentication, virtual mouse, smart wheelchair, robotic ophthalmic surgery, etc. However, iris detection is difficult due to specular reflection, occlusion, glistening on glasses, the distance between a person's eye and the camera, etc. Therefore, an integrated approach for iris detection and tracking in an uncontrolled environment is proposed. The proposed approach consists of (i) Tiny-YOLOv3-based eye detection, (ii) Seg-Net for iris detection/segmentation, and (iii) a KLT algorithm for iris tracking. Tracking reduces the computational complexity after the segmentation/localization of the iris in the initial frame instead of detection/segmentation across each frame. The models are evaluated on various benchmark databases (i) BioID, (ii) GI4E, (iii) Talking-Face, and (iv) NITSGoP databases for eye detection, GI4E and NITSGoP databases for iris segmentation, and localization. The complete (eye detection, iris segmentation, iris tracking) model is evaluated on NITSGoP and UPNA head poses databases for iris tracking. Extensive experiments show that our proposed method outperforms the baseline methods. The results also indicate that the proposed method can overcome the iris segmentation in each frame to reduce the computational complexity.
... Ahmad et al. [14] suggested Faster Recurrent Convolutional Neural Networks (FRCNN), AlexNet, and a Rectangular-Intensity-Gradient (RIG) technique as a unified method for face recognition, eye recognition, detection of eye openness, and eye center localization. Both eyes were identified using Faster RCNN, then, AlexNet assisted in determining the eye's status (closed or open). ...
... However, it is a challenging task. Face rotation, occlusion, facial expression, image quality, and wearing glasses make eye detection difficult [3,9,15,17]. There are two main approaches to eye detection (i) Non-neural-based methods and (ii) Neural-based methods. ...
... Therefore, low computation with higher accuracy is the need of the hour for eye detection. Region-based convolutional neural (RCNN) object detection motivated us to implement Faster RCNN for eye detection [9]. Therefore, a Faster RCNN deep learning model based on a region proposal network for face and eye detection was implemented on a customized dataset through an image labeler [10]. ...
... Therefore, Faster RCNN will detect face first, and then detected face will limit the search area for eye detection. Faster RCNN combines feature extractor, region proposal network (RPN), and the detector network [9]. ResNet 101 extracts deep features [11]. ...
Eye detection is essential in many computer vision
applications such as driver drowsiness detection, human behavior
analysis, liveness detection, gaze estimation, etc. However, eye
detection in a facial image is challenging due to face rotation, pose
variation, scale variation, occlusion, etc. Therefore, Faster RCNN
deep learning-based eye detection model is proposed under the
occlusion and pose variation. Pre-processing was done using
contrast stretching which makes the database better for detection.
Fine-tuning of hyperparameters and augmentation makes the
detector more accurate and robust. The effectiveness of the eye
detection model was analyzed on the publicly available AR and
GI4E databases. This model gives 98.32% and 98.11% accuracy
for the AR and GI4E databases with a 0.52 ms/image
computational time. Extensive experiments suggested that our
proposed model resulted better than state-of-the-art methods for
eye detection.
Eye centre localisation is critical to eye tracking systems of various forms and with applications in variety of disciplines. An active eye tracking approach can achieve a high accuracy by leveraging active illumination to gain an enhanced contrast of the pupil to its neighbourhood area. While this approach is commonly adopted by commercial eye trackers, a dependency on IR lights can drastically increase system complexity and cost, and can limit its range of tracking, while reducing system usability. This paper investigates into a passive eye centre localisation approach, based on a single camera, utilising convolutional neural networks. A number of model architectures were experimented with, including the Inception-v3, NASNet, MobileNetV2, and EfficientNetV2. An accuracy of 99.34% with a 0.05 normalised error was achieved on the BioID dataset, which outperformed four other state-of-the-art methods in comparison. A means to further improve this performance on high-resolution data was proposed; and it was validated on a high-resolution dataset containing 12,381 one-megapixel images. When assessed in a typical eye tracking scenario, an average eye tracking error of 0.87% was reported, comparable to that of a much more expensive commercial eye tracker.KeywordsEye centre localisationEye trackingDeep learning
