Filters








2,821 Hits in 4.1 sec

Colour Invariant Head Pose Classification in Low Resolution Video

B. Benfold, I. Reid
2008 Procedings of the British Machine Vision Conference 2008  
This paper presents an algorithm for the classification of head pose in low resolution video.  ...  Results from video sequences demonstrate that an improved posterior estimation using learnt colour distributions reduces classification error and provides accurate pose information in images where the  ...  Conclusion It has been shown that segmented head images provide enough information to allow pose estimation in low resolution video, removing the need for any assumptions about hair or skin colours.  ... 
doi:10.5244/c.22.49 dblp:conf/bmvc/BenfoldR08 fatcat:di2ze4eqnzcbldydvucr72syey

Review of person re-identification techniques

Mohammad Ali Saghafi, Halimah Badioze Zaman, Mohamad Hanif Md. Saad, Aini Hussain
2014 IET Computer Vision  
In all of the existing re-identification approaches, feature vectors are extracted from segmented still images or video frames.  ...  Person re-identification across different surveillance cameras with disjoint fields of view has become one of the most interesting and challenging subjects in the area of intelligent video surveillance  ...  employed in combination with low-resolution cameras.  ... 
doi:10.1049/iet-cvi.2013.0180 fatcat:n5y3x2ooqjeo3iarj7g7nxvpti

Real-time Full-body Visual Traits Recognition from Image Sequences [article]

Christoph Jung, Reimar Tausch, Christian Wojek
2010 International Symposium on Vision, Modeling, and Visualization  
Obviously full-body classification is more challenging, mainly due to large variations in body pose, clothing and occlusion.  ...  Few systems have recently applied recognition on low resolution full-body images, which shows lower performance than the facial regions but already can deliver classification results even if no face is  ...  All images have a relatively low resolution of 64 × 128 pixels with the person centered taking approx. 100 pixels in height (figure 4 ).  ... 
doi:10.2312/pe/vmv/vmv10/057-064 dblp:conf/vmv/JungTW10 fatcat:yxarytmyordlnjksb3xhgl45ki

Automated Person Identification in Video [chapter]

Mark Everingham, Andrew Zisserman
2004 Lecture Notes in Computer Science  
We describe progress in the automatic detection and identification of humans in video, given a minimal number of labelled faces as training data.  ...  The method we have developed combines approaches from computer vision, for detection and pose estimation, with those from machine learning for classification.  ...  We must cope with large changes in scale: faces vary in size from 200 pixels to as little as 15 pixels (i.e. very low resolution), partial occlusion, varying lighting, poor image quality, and motion blur  ... 
doi:10.1007/978-3-540-27814-6_36 fatcat:7tpocsxagrfcbjhtixwydlqzc4

Semi-supervised intelligent surveillance system for secure environments

C. Fookes, S. Denman, R. Lakemond, D. Ryan, S. Sridharan, M. Piccardi
2010 2010 IEEE International Symposium on Industrial Electronics  
Recent advances in these computer vision modules and capability gaps in surveillance technology are also highlighted.  ...  critical surveillance tasks including: the management and calibration of cameras within a multi-camera network; tracking of objects across multiple views; recognition of people utilising biometrics and in  ...  This helps improve image fidelity and recognition rates when dealing with low-resolution faces and silhouettes.  ... 
doi:10.1109/isie.2010.5636922 fatcat:hsuepfx5evbmvir5zv2726jo3e

Automatic Detection of ADHD and ASD from Expressive Behaviour in RGBD Data [article]

Shashank Jaiswal, Michel Valstar, Alinda Gillott, David Daley
2016 arXiv   pre-print
To do so, we conduct the questionnaires in a computer-mediated way while recording participants with modern RGBD (Colour+Depth) sensors.  ...  in adults.  ...  This resulted in a 54 dimensional head pose vector F hp , for each video segment. 4) Speed of head movement: Dynamics of head motion has been a less researched aspect in the field of psychological disorders  ... 
arXiv:1612.02374v1 fatcat:vozzds5htjg4hbcdfu2jtyiocq

Fully automated image-based estimation of postural point-features in children with cerebral palsy using deep learning

Ryan Cunningham, María B. Sánchez, Penelope B. Butler, Matthew J. Southgate, Ian D. Loram
2019 Royal Society Open Science  
Three expert operators manually annotated 13 point-features in every fourth image of 177 short (5-10 s) videos (25 Hz) of 12 children with cerebral palsy (aged: 4.52 ± 2.4 years), participating in SATCo  ...  Truncal segment angles (head, neck and six thoraco-lumbar-pelvic segments) were estimated with error 6.4 ± 2.8°, allowing accurate classification (F 1 > 80%) of deviation from a reference posture at thresholds  ...  The authors express sincere thanks to the children whose data were used in this study, their families and to the staff of The Movement Centre, Oswestry, UK. royalsocietypublishing.org/journal/rsos R.  ... 
doi:10.1098/rsos.191011 pmid:31827842 pmcid:PMC6894590 fatcat:uvuquqn5bnbylcnxqmsy7htrnq

Review of constraints on vision-based gesture recognition for human–computer interaction

Biplab Ketan Chakraborty, Debajit Sarma, M.K. Bhuyan, Karl F MacDorman
2018 IET Computer Vision  
Current challenges are explored in detail.  ...  The ability of computers to recognize hand gestures visually is essential for progress in human-computer interaction.  ...  Using a single colour video camera and consequently the projection of the 3D scene on a 2D plane results in the loss of depth information.  ... 
doi:10.1049/iet-cvi.2017.0052 fatcat:fs2l3o27vveqjlanldjmziuudi

Efficient CNN Implementation for Eye-Gaze Estimation on Low-Power/Low-Quality Consumer Imaging Systems [article]

Joseph Lemley, Anuradha Kar, Alexandru Drimbarean, Peter Corcoran
2018 arXiv   pre-print
Such systems are required to operate reliably in difficult, unconstrained environments with low power consumption and at minimal cost.  ...  In this paper a new hardware friendly, convolutional neural network model with minimal computational requirements is introduced and assessed for efficient appearance-based gaze estimation.  ...  Another high resolution, head-pose-invariant tracking method that does not require geometric models is presented by [19] .  ... 
arXiv:1806.10890v1 fatcat:cp6k4rx3gzc33morknlhcpw7di

Deep Head Pose: Gaze-Direction Estimation in Multimodal Video

Sankha S. Mukherjee, Neil Martin Robertson
2015 IEEE transactions on multimedia  
In this paper we present a convolutional neural network (CNN)-based model for human head pose estimation in low-resolution multi-modal RGB-D data.  ...  We present state-of-the-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  ...  Head Pose Estimation The pioneering work on low resolution head pose estimation was proposed by Robertson and Reid [4] which used a detector based on template training to classify head poses in eight  ... 
doi:10.1109/tmm.2015.2482819 fatcat:qrz7fhtnynbjbfwt6eqlr4enw4

Automatic Detection of ADHD and ASD from Expressive Behaviour in RGBD Data

Shashank Jaiswal, Michel F. Valstar, Alinda Gillott, David Daley
2017 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017)  
To do so, we conduct the questionnaires in a computer-mediated way while recording participants with modern RGBD (Colour+Depth) sensors.  ...  in adults.  ...  This resulted in a 54 dimensional head pose vector F hp , for each video segment. 4) Speed of head movement: Dynamics of head motion has been a less researched aspect in the field of psychological disorders  ... 
doi:10.1109/fg.2017.95 dblp:conf/fgr/JaiswalVGD17 fatcat:pyi257pv3jezpp2q7ihvqni6tq

Automated Face Recognition: Challenges and Solutions [chapter]

Joanna Isabelle Olszewska
2016 Pattern Recognition - Analysis and Applications  
involves important variations in both acquisition conditions as well as in facial expressions and in pose changes.  ...  Automated face recognition (AFR) aims to identify people in images or videos using pattern recognition techniques.  ...  , low resolution and out-of-focus faces.  ... 
doi:10.5772/66013 fatcat:hty4tc6i6zaq5m3fxh7stn4via

Human action recognition using an ensemble of body-part detectors

Bhaskar Chakraborty, Andrew D. Bagdanov, Jordi Gonzàlez, Xavier Roca
2011 Expert systems  
Three example-based body part detectors are trained to detect three components of the human body: the head, the legs and the arms.  ...  Grouping of these detections is then performed using a simple geometric constraint model which yields a viewpoint invariant human detector.  ...  HumanEva dataset contains 7 calibrated video sequences (4 grayscale and 3 colour) that are synchronized with 3D body poses obtained from a motion capture system.  ... 
doi:10.1111/j.1468-0394.2011.00610.x fatcat:x6ermka5pve2tih3syfoilz2fm

Head Direction Estimation from Silhouette [chapter]

Amina Bensebaa, Slimane Larabi, Neil M. Robertson
2013 Lecture Notes in Computer Science  
Due to the absence of features that may be extracted from face, heading direction estimation for low resolution images is a difficult task and requires the taking into account all information that may  ...  We propose in this paper a set of geometric features extracted from shape head-shoulders, feet and knees shapes which jointly allow the estimation of body direction.  ...  Using multiple views camera, Voit et al [17] estimate head pose for low resolution image by appearance-based method.  ... 
doi:10.1007/978-3-642-41181-6_35 fatcat:nnoonfhbqvfljedxlooqnsueva

Pre-Attentive Face Detection for Foveated Wide-Field Surveillance

S.J.D. Prince, J.H. Elder, Y. Hou, M. Sizinstev
2005 2005 Seventh IEEE Workshops on Applications of Computer Vision (WACV/MOTION'05) - Volume 1  
This problem may be overcome by combining a fixed, preattentive, low-resolution wide-field camera with a shiftable, attentive, high-resolution narrow-field camera.  ...  Unfortunately, the low image resolution of the widefield sensor precludes the use of most conventional facedetection algorithms.  ...  We present a robust method for locating potential head regions in very low resolution images. This technique combines three weak cues: skin colour, motion, and foreground extraction.  ... 
doi:10.1109/acvmot.2005.95 dblp:conf/wacv/PrinceEHS05 fatcat:5t34jhjdmzaqdcu4chrtmcwita
« Previous Showing results 1 — 15 out of 2,821 results