Detecting semantic events in sports video is crucial for video indexing and retrieval. Most existing works have exclusively relied on video content features, namely, directly available and extractable data from the visual and/or aural channels. Sole reliance on such data however, can be problematic due to the high-level semantic nature of video and the difficulty to properly align detected events with their exact time of occurrences. This paper proposes a framework for soccer goal event detection through collaborative analysis of multimodal features. Unlike previous approaches, the visual and aural contents are not directly scrutinized. Instead, an external textual source (i.e., minute-by-minute reports from sports websites) is used to initially localize the event search space. This step is vital as the event search space can significantly be reduced. This also makes further visual and aural analysis more efficient since excessive and unnecessary non-eventful segments are discarded, culminating in the accurate identification of the actual goal event segment. Experiments conducted on thirteen soccer matches are very promising with high accuracy rates being reported.
To segment an image using the random walks algorithm; users are often required to initialize the approximate locations of the objects and background in the image. Due to its segmenting model that is mainly reflected by the relationship among the neighborhood pixels and its boundary conditions, random walks algorithm has made itself sensitive to the inputs of the seeds. Instead of considering the relationship between the neighborhood pixels solely, an attempt has been made to modify the weighting function that accounts for the intensity changes between the neighborhood nodes. Local affiliation within the defined neighborhood region of the two nodes is taken into consideration by incorporating an extra penalty term into the weighting function. Besides that, to better segment images, particularly medical images with texture features, GLCM variance is incorporated into the weighting function through kernel density estimation (KDE). The probability density of each pixel belonging to the initialized seeds is estimated and integrated into the weighting function. To test the performance of the proposed weighting model, several medical images that mainly made up of 174-brain tumor images are experimented. These experiments establish that the proposed method produces better segmentation results than the original random walks.
Automating the detection of lesions in liver CT scans requires a high performance and robust solution. With CT-scan start to become the norm in emergency department, the need for a fast and efficient liver lesions detection method is arising. In this paper, we propose a fast and evolvable method to profile the features of pre-segmented healthy liver and use it to detect the presence of liver lesions in emergency scenario. Our preliminary experiment with the MICCAI 2007 grand challenge datasets shows promising results of a fast training time, ability to evolve the produced healthy liver profiles, and accurate detection of the liver lesions. Lastly, the future work directions are also presented.
This paper presents two novel approaches to determine optimum growing multi-experts network (GMN) structure. The first method called direct method deals with expertise domain and levels in connection with local experts. The growing neural gas (GNG) algorithm is used to cluster the local experts. The concept of error distribution is used to apportion error among the local experts. After reaching the specified size of the network, redundant experts removal algorithm is invoked to prune the size of the network based on the ranking of the experts. However, GMN is not ergonomic due to too many network control parameters. Therefore, a self-regulating GMN (SGMN) algorithm is proposed. SGMN adopts self-adaptive learning rates for gradient-descent learning rules. In addition, SGMN adopts a more rigorous clustering method called fully self-organized simplified adaptive resonance theory in a modified form. Experimental results show SGMN obtains comparative or even better performance than GMN in four benchmark examples, with reduced sensitivity to learning parameters setting. Moreover, both GMN and SGMN outperform the other neural networks and statistical models. The efficacy of SGMN is further justified in three industrial applications and a control problem. It provides consistent results besides holding out a profound potential and promise for building a novel type of nonlinear model consisting of several local linear models.
This paper introduces an approach to perform segmentation of regions in computed tomography (CT) images that exhibit intra-region intensity variations and at the same time have similar intensity distributions with surrounding/adjacent regions. In this work, we adapt a feature computed from wavelet transform called wavelet energy to represent the region information. The wavelet energy is embedded into a level set model to formulate the segmentation model called wavelet energy-guided level set-based active contour (WELSAC). The WELSAC model is evaluated using several synthetic and CT images focusing on tumour cases, which contain regions demonstrating the characteristics of intra-region intensity variations and having high similarity in intensity distributions with the adjacent regions. The obtained results show that the proposed WELSAC model is able to segment regions of interest in close correspondence with the manual delineation provided by the medical experts and to provide a solution for tumour detection.