1,721,209 research outputs found
Learning in Markov Games: can we exploit a general-sum opponent?
No full textIn this paper, we study the learning problem in two-player general-sum Markov Games. We consider the online setting where we control a single player, playing against an arbitrary opponent to minimize the regret. Previous works only consider the zero-sum Markov Games setting, in which the two agents are completely adversarial. However, in some cases, the two agents may have different reward functions without having conflicting objectives. This involves a stronger notion of regret than the one used in previous works. This class of games, called general-sum Markov Games is far to be well understood and studied. We show that the new regret minimization problem is significantly harder than in standard Markov Decision Processes and zero-sum Markov Games. To do this, we derive a lower bound on the expected regret of any “good” learning strategy which shows the constant dependencies with the number of deterministic policies, which is not present in zero-sum Markov Games and Markov Decision Processes. Then we propose a novel optimistic algorithm that nearly matches the proposed lower bound. Proving these results requires overcoming several new challenges that are not present in Markov Decision Processes or zero-sum Markov Games
Speeded-up Convolution Neural Network for classification tasks using multiscale 2-dimensional decomposition
Full text linkIn this paper, we propose a strategy for network simplification and acceleration. First, we propose to generate a suitable resized image using multiscale patching in the first convolutional layer, which can then be used for the rest of the network. We use p convolutional filters that operate on patches of size m × n, and we first select all the possible non-superposed m × n patches from the available images. If the number of such patches is not sufficient, the remaining ones are collected using scales a × b or c × d such that ab = cd = mn or a = 2 m or b = 2n. Patches generated from the former condition are directly extracted from the images, while the downsampled results are used in the latter case. We also introduce a 2-dimensional decomposition for patch compression, by stacking all the available image patches along the columns and applying a 2D PCA decomposition. Finally, a layer weight decomposition technique followed by module-based finetuning is adopted, for a new fast module-based CNN model. Extensive evaluations using public data sets like MNIST, Pascal VOC, and WebCASIA, and with different state-of-the art CNN architectures like Open-Face, VGG and DarkNet verify that our proposed model is able to accelerate the training process and even to provide higher classification accuracies for small-sized datasets. We obtain 8% increase in Top-1 and Top-5 recognition rates, and 5% increase in F1 score over general interpolation-based resizing
Nonlinear kernel based feature maps for blur-sensitive unsharp masking of JPEG images
Full text linkIn this paper, a method for estimating the blur regions of an image is first proposed, resorting to a mixture of linear and nonlinear convolutional kernels. The blur map obtained is then utilized to enhance images such that the enhancement strength is an inverse function of the amount of measured blur. The blur map can also be used for tasks such as attention-based object classification, low light image enhancement, and more. A CNN architecture is trained with nonlinear upsampling layers using a standard blur detection benchmark dataset, with the help of blur target maps. Further, it is proposed to use the same architecture to build maps of areas affected by the typical JPEG artifacts, ringing and blockiness. The blur map and the artifact map pair permit to build an activation map for the enhancement of a (possibly JPEG compressed) image. Extensive experiments on standard test images verify the quality of the maps obtained using the algorithm and their effectiveness in locally controlling the enhancement, for superior perceptual quality. Last but not least, the computation time for generating these maps is much lower than the one of other comparable algorithms
Virtual Restoration of Faded Photographic Prints
No full textAntique photographic prints are subject to fading due to the action of time and of diverse chemical agents. A method for automated virtual restoration of digital images obtained from scanned photographic prints is proposed in this paper. The effects of film grain noise are also taken into consideration. Experimental results on archive photographic material show the performances of the proposed technique. 1
Algorithmic and architectural design for real-time and power-efficient Retinex image/video processing
No full textThis paper presents novel algorithmic and architectural solutions for real-time and power-efficient enhancement of images and video sequences. A programmable class of Retinex-like filters, based on the separation of the illumination and reflectance components, is proposed. The dynamic range of the input image is controlled by applying a suitable non-linear function to the illumination, while the details are enhanced by processing the reflectance. An innovative spatially recursive rational filter is used to estimate the illumination. Moreover, to improve the visual quality results of two-branch Retinex operators when applied to videos, a novel three-branch technique is proposed which exploits both spatial and temporal filtering. Real-time implementation is obtained by designing an Application Specific Instruction-set Processor (ASIP). Optimizations are addressed at algorithmic and architectural levels. The former involves arithmetic accuracy definition and linearization of non-linear operators; the latter includes customized instruction set, dedicated memory structure, adapted pipeline, bypasses, custom address generator, and special looping structures. The ASIP is synthesized in standard-cells CMOS technology and its performances are compared to known Digital signal processor (DSP) implementations of real-time Retinex filters. As a result of the comparison, the proposed algorithmic/architectural design outperforms state-of-art Retinex-like operators achieving the best trade-off between power consumption, flexibility, and visual quality
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Whole field optical profilometry: application of nonlinear processing algorithms to the enhancement of low-contrast images
No full textDesign of ASIP (Application Specific Instruction-set Processor) architectures for image processing in CMOS VLSI technologies
No full textOBJECTIVE MEASURES FOR THE EVALUATION OF TECHNIQUES FOR THE VIRTUAL RESTORATION OF FADED SEPIA PHOTOGRAPHIC PRINTS
No full text- …
