26 research outputs found

    Single image deblurring with or without face prior and its applications

    No full text
    The motion blur is one of the most difficult challenges in photography, which is generated from the relative motion between the sensor and the scene during exposure time. These blur artifacts degrade the visual experience, and the performance of various applications, such as, object detection, facial analysis. Therefore, it is significant to remove the blur and restore sharp and clean images. Our work focuses on the general single image deblurring, and face image deblurring with face prior. State-of-the-art single image deblurring techniques are sensitive to image noise. Even a small amount of noise, which is inevitable in low-light conditions, can degrade the quality of blur kernel estimation dramatically. We propose a new method for handling noise in blind image deconvolution based on new theoretical and practical insights. Based on the observations on directional filter, our method applies a series of directional filters at different orientations to the input image, and estimates an accurate Radon transform of the blur kernel from each filtered image. Finally, we reconstruct the blur kernel using inverse Radon transform. Experimental results on synthetic and real data show that our algorithm achieves higher quality results than previous approaches on blurry and noisy images. The human face is one of the most essential focuses in numerous applications. Although significant progress has been made in the image deblurring area, few of them can obtain promising results on blurry face images. Many state-of-the-art single image deblurring approaches estimate the blur kernel based on analyzing the edge profiles of the input image. However, the detection of strong edges is very difficult on human faces, since the human faces do not contain as much texture as natural images. We propose to utilize the global face structure information to help with the strong or salient edge detection. Our method outperforms the existing methods in extensive evaluations on synthetic and real face images. Facial expression is a significant application on sharp and restored face images. To improve the general facial expression recognition performance, we present a new idea to analyze facial expression by exploring the common and specific information among different expressions.Ph.D.Includes bibliographical referencesby Lin Zhon

    Memory, language and trauma in the work of Félix Grande

    No full text
    PhDMy thesis explores how memory and trauma permeate the work of the poet Félix Grande (Mérida, Spain, 1937). It addresses the question of how his particular understanding of memory is opposed to a rather bleak view of it held by many other Spanish poets of the time. Grande does not yield to a generalized discrediting of memory. On the contrary, memory is the driving force behind his writing, and this thesis constitutes an analysis of its mechanisms. The originality of Grande’s work stems from the ways in which it shares common ground with contemporary research carried out by disciplines that integrate Memory and Trauma Studies. His poetic voice struggles to grasp aspects of memory whose articulation proves traumatic. These elements resist symbolic translation and turn his poetry into a work of constant rumination without closure. Grande’s work illustrates that literature is both inextricably linked to memory, and is well equipped to deal with trauma, as the labour carried out by memory, weaving and un-weaving, especially in its attempts to mourn, is at the heart of his artistic production. Finally, his work instantiates a relationship with language and memory which, while recognising the limits of language to express and of memory to retrieve the past, goes beyond this initial distrust to offer a positive perspective on these faculties, as the means for establishing modes of survival and rethinking our connections to the unknown
    corecore