Transfer learning allows leveraging the knowledge of source domains, available a priori, to help training a classifier for a target domain, where the available data is scarce. The effectiveness of the transfer is affe...
详细信息
Transfer learning allows leveraging the knowledge of source domains, available a priori, to help training a classifier for a target domain, where the available data is scarce. The effectiveness of the transfer is affected by the relationship between source and target. Rather than improving the learning, brute force leveraging of a source poorly related to the target may decrease the classifier performance. One strategy to reduce this negative transfer is to import knowledge from multiple sources to increase the chance of finding one source closely related to the target. This work extends the boosting framework for transferring knowledge from multiple sources. Two new algorithms, MultiSource-TrAdaBoost, and TaskTrAdaBoost, are introduced, analyzed, and applied for object category recognition and specific object detection. The experiments demonstrate their improved performance by greatly reducing the negative transfer as the number of sources increases. TaskTrAdaBoost is a fast algorithm enabling rapid retraining over new targets.
We present a surface radiance model for diffuse lighting that incorporates shadows, interreflections, and surface orientation. We show that, for smooth surfaces, the model is an excellent approximation of the radiosit...
详细信息
We present a surface radiance model for diffuse lighting that incorporates shadows, interreflections, and surface orientation. We show that, for smooth surfaces, the model is an excellent approximation of the radiosity equation. We present a new data structure and algorithm that uses this model to compute shape-from-shading under diffuse lighting. The algorithm was tested on both synthetic and real images, and performs more accurately than the only previous algorithm for this problem. Various causes of error are discussed, including approximation errors in image modelling, poor local constraints at the image boundary, and ill-conditioning of the problem itself.
A hierarchical stereo system is described that uses structural descriptions up to the surface level. Surface descriptions are computed from monocular images, by using a perceptual grouping technique. Occlusion can be ...
详细信息
A hierarchical stereo system is described that uses structural descriptions up to the surface level. Surface descriptions are computed from monocular images, by using a perceptual grouping technique. Occlusion can be a major problem in stereo analysis and is often not treated explicitly. An analysis is presented of occlusion effects in stereo, and it is shown how structural descriptions can be used to deal with them. Experimental results are given for scenes with curved objects and significant occlusions.< >
A novel scene reconstruction technique is presented, different from previous approaches in its ability to cope with large changes in visibility and its modeling of intrinsic scene color and texture information. The me...
详细信息
A novel scene reconstruction technique is presented, different from previous approaches in its ability to cope with large changes in visibility and its modeling of intrinsic scene color and texture information. The method avoids image correspondence problems by working in a discretized scene space whose voxels are traversed in a fixed visibility ordering. This strategy takes full account of occlusions and allows the input cameras to be far apart and widely distributed about the environment. The algorithm identifies a special set of invariant voxels which together form a spatial and photometric reconstruction of the scene, fully consistent with the input images. The approach is evaluated with images from both inward- and outward-facing cameras.
Subject covariate data were collected on 1, 072 pairs of FERET images for analysis in a human face verification experiment. The subject data included information about facial hair, bangs, eyes, gender, and age. The ve...
详细信息
Subject covariate data were collected on 1, 072 pairs of FERET images for analysis in a human face verification experiment. The subject data included information about facial hair, bangs, eyes, gender, and age. The verification experiment was replicated at seven different false alarm rates ranging from 1/10, 000 to 1/100. A generalized linear mixed model (GLMM) was fit to the binary outcomes indicating correct verification. Statistically significant main effects for bangs, eyes, gender, and age were found. The effect of the log false positive rate on verification success was found to interact significantly with bangs, gender, and age. These results have important implications for future evaluation of biometrics, and the GLMM methodology used here is shown to be effective and informative for this sort of data.
Previous algorithms that recover camera motion from image velocities suffer from both bias and excessive variance in the results. We propose a robust estimator of camera motion that is statistically consistent when im...
详细信息
Previous algorithms that recover camera motion from image velocities suffer from both bias and excessive variance in the results. We propose a robust estimator of camera motion that is statistically consistent when image noise is isotropic. Consistency means that the estimated motion converges in probability, to the true value as the number of image points increases. An algorithm based on reweighted Gauss-Newton iterations handles 100 velocity measurements in about 50 milliseconds on a workstation.
A novel depth computation algorithm is described. Based on linear equations derived from the consideration of relative normal velocity, this algorithm determines depth values locally and uniquely from normal velocitie...
详细信息
A novel depth computation algorithm is described. Based on linear equations derived from the consideration of relative normal velocity, this algorithm determines depth values locally and uniquely from normal velocities. Multiple independently moving objects present no problem for the algorithm. A velocity constraint is implemented to estimate normal velocities from intensity images. Experiments to determine depth values from both synthetic images and real images are presented. The experiments show that the depth-recovery algorithm can recover correct values from intensity images. The accuracy of the depth values depends heavily on the accuracy of the normal velocities.< >
We advocate the use of Gaussian Process Dynamical Models (GPDMs) for learning human pose and motion priors for 3D people tracking. A GPDM provides a lowdimensional embedding of human motion data, with a density functi...
详细信息
We advocate the use of Gaussian Process Dynamical Models (GPDMs) for learning human pose and motion priors for 3D people tracking. A GPDM provides a lowdimensional embedding of human motion data, with a density function that gives higher probability to poses and motions close to the training data. With Bayesian model averaging a GPDM can be learned from relatively small amounts of data, and it generalizes gracefully to motions outside the training set. Here we modify the GPDM to permit learning from motions with significant stylistic variation. The resulting priors are effective for tracking a range of human walking styles, despite weak and noisy image measurements and significant occlusions.
Occlusion is usually modelled in two images symmetrically in previous stereo algorithms which cannot work for multi-view stereo efficiently. In this paper, we present a novel formulation that handles occlusion using o...
详细信息
ISBN:
(纸本)0769523722
Occlusion is usually modelled in two images symmetrically in previous stereo algorithms which cannot work for multi-view stereo efficiently. In this paper, we present a novel formulation that handles occlusion using only one depth map in an asymmetrical way. Consequently, multi-view information is efficiently accumulated to achieve high accuracy. The resulting energy function is complex and approximate graph cut based solutions are proposed. Our approach complements the theory and extends the applicability of using graph cut in stereo. The experiments demonstrate that the approach is comparable with the state of the art and potentially more efficient for multi-view stereo.
In this paper, to evaluate the performance of object recognition algorithms, we propose a new evaluation framework by synthesizing natural scenes based on the Amsterdam Library of Object Images [1]. Here, the evaluati...
详细信息
In this paper, to evaluate the performance of object recognition algorithms, we propose a new evaluation framework by synthesizing natural scenes based on the Amsterdam Library of Object Images [1]. Here, the evaluation of an object recognition algorithm has the basis on searching an object in a synthetic scene. More specifically, an object is selected, and then the synthetic scene under a specific condition is generated by using images, affected by that condition, of that object and other objects in the database. When generating the synthetic scene, the other objects are randomly selected and all objects are naturally distributed in the synthetic scene. Let us refer to this synthetic scene as a Virtual Scene. Then the performance of object recognition algorithms for the specific condition can be analyzed by using a group of Virtual Scenes in that condition. As an example of utilizing the proposed framework, an object recognition algorithm using the scale-invariant feature transform [2] has been evaluated and analyzed in the case of changing the viewing direction, illumination color, and illumination direction.
暂无评论