Showing papers in "Image and Vision Computing in 2013"

TL;DR: This survey paper aims to put the continuity aspect of affect under the spotlight by investigating the current trends and providing guidance towards possible future directions.

TL;DR: Comparing the results with the recognition scores of all Audiovisual Sub-Challenge participants, it is found that the proposed LSTM-based technique leads to the best average recognition performance that has been reported for this task so far.

TL;DR: This work proposes to involve the user and investigate methods for implicit tagging, wherein users' responses to the interaction with the multimedia content are analyzed in order to generate descriptive tags.

TL;DR: A new edge detector based on an edge and acquisition model derived from the partial area effect, which does not assume continuity in the image values is presented, achieving a highly accurate extraction of the position, orientation, curvature and contrast of the edges.

TL;DR: A Gaussian Mixture Model-based approach is applied which computes a mapping from a set of observed audio-visual cues to an underlying emotional state and sheds light into the way expressive body language is modulated by underlying emotional states in the context of dyadic interactions.

TL;DR: A new publicly available audiovisual database suitable for studying laughter is presented, the MAHNOB Laughter database, and the combination of audio and visual information is beneficial in the presence of acoustic noise and helps discriminating between voiced laughter episodes and speech utterances.

TL;DR: This work develops a computational method to predict the BMI from face images automatically, formulate the BMI prediction from facial features as a machine vision problem, and evaluates the approach on a large database with more than 14,500 face images.

TL;DR: The nonverbal behavioural cues that could be present during displays of agreement and disagreement are surveyed; a number of methods that can be used or adapted to detect these suggested cues are discussed; some publicly available databases these tools could be trained on for the analysis of spontaneous, audiovisual instances of agreement or disagreement are listed.

TL;DR: An efficient method for text-independent writer identification using a codebook method that uses the occurrence histogram of the shapes in a code book to create a feature vector for each specific manuscript is proposed.

TL;DR: The state of the art of such acquisition systems and their applications for the study of cultural heritage are presented and existing registration techniques that can be used to obtain 3D models with multispectral texture are described.

TL;DR: The proposed panoramic gait recognition framework can greatly reduce the complexity of the classification problem while achieving fair correct classification rates when gait is captured with unknown conditions.

TL;DR: A new level set method integrating local and global intensity information adaptively to segment inhomogeneous images to avoid the re-initialization of the level set function and shorten the computational time is proposed.

TL;DR: A novel forgery detection algorithm to recognize tampered inpainting images, which is one of the effective approaches for image manipulation, and a two-stage searching algorithm based on weight transformation to speed up the computation speed is proposed.

TL;DR: An original method to extract the 3D head trajectory of a person in a room is proposed using only one calibrated camera, represented as a 3D ellipsoid, which is tracked with a hierarchical particle filter based on color histograms and shape information.

TL;DR: This paper combines both local motion and appearance feature in a novel framework to model the temporal dynamics of face and body gesture and proposes a bag of words (BOW) based representation for both MHI-HOG and Image- HOG features.

TL;DR: Experimental results obtained with different challenging datasets show that the proposed system exceeds state-of-the-art performance, both in terms of localization and recognition.

TL;DR: This paper presents an automatic wildfire smoke detection method using computer vision and pattern recognition techniques that can increase the detection accuracy performance when compared with related methods and allow smoke detection to be carried out in near real time.

TL;DR: The concept and applications of Nonverbal Communication Computing are introduced and some of the motion analysis methods employed in this area are reviewed, which include face tracking, expression recognition, body reconstruction, and group activity analysis.

TL;DR: Experiments show the non-parametric method for estimating the pose of human hands to outperform other state of the art regression methods, while operating at a significantly lower computational cost than comparable model-based hand tracking methods.

TL;DR: Experimental results show that this new inpainting approach based on the fast marching method performs better than the local methods in terms of both visual and metric qualities, and it achieves visually comparable results to the time-consuming global method.

TL;DR: The proposed method is based on the bag of video words (BOV) representation and does not require prior knowledge about actions, background subtraction, motion estimation or tracking, and is robust to spatial and temporal scale changes, as well as some deformations.

TL;DR: This paper proposes an On-line Appearance-Based Tracker for simultaneous tracking of 3D head pose, lips, eyebrows, eyelids and irises in monocular video sequences using a hierarchical combination of three OABTs optimized using a Levenberg-Marquardt Algorithm enhanced with line-search procedures.

TL;DR: G-SURF descriptors are fast to compute, but have extra matching robustness due to the extra invariance offered by gauge derivatives, and extensive experimental image matching results on the Mikolajczyk and Schmid dataset are presented.

TL;DR: A comparative study of two approaches for road traffic density estimation using the microscopic parameters which are extracted using both motion detection and tracking methods from a video sequence based on their robustness to the classification of different road traffic states.

TL;DR: A novel background modeling that is applicable to any spatio-temporal non-parametric moving object detection strategy is proposed and significantly improves the quality of previous strategies while maintaining the computational requirements of the detection process.

TL;DR: An algorithm for navigating a mobile robot using the visual potential, which dynamically selects a local pathway to the destination without collision with obstacles and without any knowledge of the robot workspace, is developed.

TL;DR: An improved sparse paradigm is introduced and state-of-the-art classification accuracy is demonstrated for the facial understanding problems of expression, gender, race, glasses, and facial hair classification.

TL;DR: A unified probabilistic framework based on a novel Bayesian Belief Network (BBN) for 3D facial expression and Action Unit (AU) recognition and its robustness in landmark localization errors is presented.

TL;DR: A novel technique for fixation prediction and saccade generation and the proposed model simulates saccadic eye movement to incorporate the underlying eye movement mechanism into saliency estimation and a simple salience measure is introduced.

TL;DR: The outcomes of this study suggest that the design and development of automated ear-based recognition systems that can operate efficiently in the lower part of the passive IR spectrum are very challenging tasks.