One of the main challenges in controlling the spread of COVID19 pandemic is to diagnose infection early. The most reliable method RT - PCR takes several hours to give results. Although the Anti-Body (Serological) test...
详细信息
Structural analysis helps in parsing the mathematical expressions. Various approaches for structural analysis have been reported in literature, but they mainly deal with online and printed expressions. In this work, t...
详细信息
ISBN:
(纸本)9789813292918;9789813292901
Structural analysis helps in parsing the mathematical expressions. Various approaches for structural analysis have been reported in literature, but they mainly deal with online and printed expressions. In this work, two-dimensional, stochastic context-free grammar is used for the structural analysis of offline handwritten mathematical expressions in a document image. The spatial relation between characters in an expression has been incorporated so that the structural variability in handwritten expressions can be tackled.
Low-light imaging using a hand-held camera is a challenging task due to excessive noise in the scene. Most of the existing methods try to address this problem either by denoising an image captured using high ISO or by...
详细信息
In this paper, a robust image watermarking scheme based on all phase sine biorthogonal transform (APSBT), singular value decomposition (SVD) and dynamic stochastic resonance (DSR) is presented. Firstly, the cover imag...
详细信息
ISBN:
(纸本)9789813292918;9789813292901
In this paper, a robust image watermarking scheme based on all phase sine biorthogonal transform (APSBT), singular value decomposition (SVD) and dynamic stochastic resonance (DSR) is presented. Firstly, the cover image is transformed by APSBT and then a gray-scale logo is embedded through the singular value decomposition. After the authentication process which essentially resolves the false-positive extraction of SVD in watermarking, a phenomenon based on dynamic stochastic resonance is deployed for the logo extraction from the watermarked image. The simulation results demonstrate that the proposed scheme has better performance in the aspect of robustness and invisibility.
In this paper, a new approach for zero-watermarking has been proposed based on log-polar mapping, all phase biorthogonal sine transform and singular value decomposition. The core idea is to produce a zero watermark to...
详细信息
ISBN:
(纸本)9789813290884;9789813290877
In this paper, a new approach for zero-watermarking has been proposed based on log-polar mapping, all phase biorthogonal sine transform and singular value decomposition. The core idea is to produce a zero watermark to protect the copyright of the image. For this purpose, the host image is transformed into the log-polar domain followed by the all phase biorthogonal sine transform (APBST). The transformed coefficients are then divided into nonoverlapping blocks and some blocks are selected based on secret key. These blocks are finally used to formulate a reference matrix which is utilized to generate a zero watermark for the host image. A detailed experimental analysis is conducted to demonstrate the feasibility of the proposed algorithm against various image/signal processing distortions.
Co-Detection is an important problem in computervision, which involves detecting common objects from multiple images. In this paper, we address the co-detection problem and propose an integrated deep learning model i...
详细信息
ISBN:
(纸本)9789813292918;9789813292901
Co-Detection is an important problem in computervision, which involves detecting common objects from multiple images. In this paper, we address the co-detection problem and propose an integrated deep learning model involving two networks for co-detection. Our proposed model detects the objects of individual images using a convolutional neural network by generating the saliency maps, which are passed as input in a Siamese neural network to ascertain whether the salient objects in both the images are similar or different. We have tested our model on the iCoseg dataset achieving high-quality results.
This paper proposes a video editor based on OpenShot with several state-of-the-art facial video editing algorithms as added functionalities. Our editor provides an easy-to-use interface to apply modern lip-syncing alg...
详细信息
The proceedings contain 80 papers. The topics discussed include: PredGAN - a deep multi-scale video prediction framework for detecting anomalies in videos;multiple kernel fisher discriminant metric learning for person...
ISBN:
(纸本)9781450366151
The proceedings contain 80 papers. The topics discussed include: PredGAN - a deep multi-scale video prediction framework for detecting anomalies in videos;multiple kernel fisher discriminant metric learning for person re-identification;vision-based steering angle prediction by the fusion of depth and intensity deep features;zero-shot learning using graph regularized latent discriminative cross-domain triplets;perfectly secure Shamir’s secret sharing scheme for privacy preserving imageprocessing over cloud;moving average recurrent neural network model for video-based person re-identification;activity recognition in egocentric videos using bag of key action units;and learning end-to-end autonomous driving using guided auxiliary supervision.
Down syndrome is a genetic disorder that affects 1 in every 1000 babies born worldwide. The cases of Down syndrome have increased in the past decade. It has been observed that humans with Down syndrome generally tend ...
详细信息
ISBN:
(纸本)9789813290884;9789813290877
Down syndrome is a genetic disorder that affects 1 in every 1000 babies born worldwide. The cases of Down syndrome have increased in the past decade. It has been observed that humans with Down syndrome generally tend to have distinct facial features. This paper proposes a model to identify people suffering from Down syndrome based on their facial features. Deep representation from different parts of the face is extracted and combined with the aid of Deep Convolutional Neural Networks. The combined representations are then classified using a Random Forest-based pipeline. The model was tested on a dataset of over 800 individuals suffering from Down syndrome and was able to achieve a recognition rate of 98.47%.
暂无评论