We propose a video copydetection scheme that employs a transform domain global video fingerprinting method. Video fingerprinting has been performed by the subspace learning based on nonnegative matrix factorization (...
详细信息
We propose a video copydetection scheme that employs a transform domain global video fingerprinting method. Video fingerprinting has been performed by the subspace learning based on nonnegative matrix factorization (NMF). It is shown that the binary video fingerprints extracted from the basis and gain matrices of the NMF representation enable us to efficiently represent the spatial and temporal content of a video segment respectively. An extensive performance evaluation has been carried out on the query and reference dataset of CBCD task of TRECVID 2011. Our results are compared with the average and the best performance reported for the task. Also NDCR and F1 rates are reported in comparison to the performance achieved via the global methods designed by the TRECVID 2011 participants. Results demonstrate that the proposed method achieves higher correct detection rates with good localization capability for the transformation of text/logo insertion, strong re-encoding, frame dropping, noise addition, gamma change or their mixtures;however there is still potential for improvement to detect copies with picture-in-picture transformations. It is also concluded that the introduced binary fingerprinting scheme is superior to the existing transform based methods in terms of the compactness.
This paper presents a novel audio fingerprinting method that is highly robust to a variety of audio distortions. It is based on unconventional audio fingerprints generation scheme. The robustness is achieved by genera...
详细信息
ISBN:
(纸本)9781479939909
This paper presents a novel audio fingerprinting method that is highly robust to a variety of audio distortions. It is based on unconventional audio fingerprints generation scheme. The robustness is achieved by generating different versions of the spectrogram matrix of the audio signal by using a threshold based on the average of the spectral values to prune this matrix. We transform each version of this pruned spectrogram matrix into a 2-D binary image. Multiple 2-D images suppress noise to a varying degree. This varying degree of noise suppression improves likelihood of one of the images matching a reference image. To speed up matching, we convert each image into an n-dimensional vector, and perform a nearest neighbor search based on this n-dimensional vector. We test this method on TRECVID 2010 content-basedcopydetection evaluation dataset. Experimental results show the effectiveness of such fingerprints even when the audio is distorted. We compare the proposed method to a state-of-the-art audio copydetection system. Results of this comparison show that our method achieves an improvement of 22% in localization accuracy, and lowers minimal normalized detection cost rate (min NDCR) by half for audio transformations T1 and T2.
Among the existing hashing methods, the Self-taught hashing (STH) is regarded as the state-of-the-art work. However, it still suffers the problem of semantic loss, which mainly comes from the fact that the original op...
详细信息
Among the existing hashing methods, the Self-taught hashing (STH) is regarded as the state-of-the-art work. However, it still suffers the problem of semantic loss, which mainly comes from the fact that the original optimization objective of in-sample data is NP-hard and therefore is compromised into the combination of Laplacian Eigenmaps (LE) and binarization. Obviously, the shape associated with the embedding of LE is quite dissimilar to that of binary code. As a result, binarization of the LE embedding readily leads to significant semantic loss. To overcome this drawback, we combine the constrained nonnegative sparse coding and the Support Vector Machine (SVM) to propose a new hashing method, called nonnegative sparse coding induced hashing (NSCIH). Here, nonnegative sparse coding is exploited for seeking a better intermediate representation, which can make sure that the binarization can be smoothly conducted. In addition, we build an image copydetection scheme based on the proposed hashing methods. The extensive experiments show that the NSCIH is superior to the state-of-the-art hashing methods. At the same time, this copydetection scheme can be used for performing copydetection over very large image database. (C) 2012 Elsevier B.V. All rights reserved.
Currently, researches on contentbased image copydetection mainly focus on robust feature extraction. However, most of existing approaches use only a single feature to represent an image for copydetection, which is ...
详细信息
Currently, researches on contentbased image copydetection mainly focus on robust feature extraction. However, most of existing approaches use only a single feature to represent an image for copydetection, which is often insufficient to characterize the image content. Besides, with the exponential growth of online images, it's urgent to explore a way of tackling the problem of large scale. In this paper, we propose a feature fusion based hashing method which effectively utilize the correlation between two feature models and efficiently accomplish large scale image copydetection. To accurately map images into the Hamming space, our hashing method not only preserves the local structure of individual feature but also globally consider the local structures for all the features to learn a group of hash functions. The experiment results show that the proposed method outperforms the state-of-the-art techniques in both accuracy and efficiency.
In this paper, a robust video copydetection system is proposed for the broadcast TV stream that is a challenging task in the wild. The proposed system extracts the signatures of video frames with Speeded-Up Robust Fe...
详细信息
ISBN:
(纸本)9781509041657
In this paper, a robust video copydetection system is proposed for the broadcast TV stream that is a challenging task in the wild. The proposed system extracts the signatures of video frames with Speeded-Up Robust Features (SURF) keypoints, which are described by Oriented Fast and Rotated Brief (ORB) descriptors for compact and efficient representation, in order to match the parts of the given reference video with the query video in interest. The resulting signal representing the similarity scores of the reference video parts with the query is then adaptively thresholded by Constant False Alarm Rate (CFAR) approach, which improves the video copydetection accuracy by exposing the strong peaks (i.e., copy locations) while discarding weak peaks/no peaks (i.e., non-copy locations), with dynamic setting of the false alarm probability. Extensive experiments on the publicly available MUSCLE-VCD-2007 Dataset and the BILGEM Video Broadcast Dataset 2015 (BVBD 2015) show that the proposed system achieves precision values of 100.0% and 91.8%, recall values of 90.0% and 99.3% respectively by proving the robustness of the system for the copydetection task in both broadcast and web video streams.
Image hashing is one of the emergent novel approaches used extensively in the field of image forensics apart from finding its place in many of the latest techniques of the area of image indexing, image retrieval etc. ...
详细信息
ISBN:
(纸本)9781509053841
Image hashing is one of the emergent novel approaches used extensively in the field of image forensics apart from finding its place in many of the latest techniques of the area of image indexing, image retrieval etc. Image hashing is basically used to identify the duplicate copies of the original images. Most of the image hashing algorithms has their limitations in getting the desirable performance against a particular image processing attack i.e. rotation. In this paper, we have proposed image hashing technique dominantly based on statistical features of the image which is robust to almost all kind of image processing attacks including rotation. In our proposed algorithm input image is normalized by using resizing, Gaussian filtering, color space conversion from RGB image to YCbCr and only Y component is taken for hash generation. Radon transform is then applied to the preprocessed image to produce 2-D Radon coefficients. 1-D DCT is then applied to the Radon coefficients to produce column-wise DCT coefficients. Lastly first AC coefficient from each column are taken to form the row-wise vector which is used to extract four statistical features, Mean, Standard Deviation, Kurtosis & Skewness. The extracted features form the final feature vector which is used for image identification. Many experiments have conducted to compare the proposed technique with the state-of-the-art techniques and the results shows that proposed hashing is robust to normal digital operations apart from giving excellent result against rotation.
content-basedcopydetection(CBCD) is a heated research area.A variety of features are extracted and applied extensively to improve the performance of CBCD systems.A novel scheme which adopts 3-Dimensional SIFT Descri...
详细信息
content-basedcopydetection(CBCD) is a heated research area.A variety of features are extracted and applied extensively to improve the performance of CBCD systems.A novel scheme which adopts 3-Dimensional SIFT Descriptors(Paul Scavenger, 2007) based on the innovative dynamic tunnels is proposed in this *** algorithm is aim to reach a proper balance between efficiency and *** results indicate that the algorithm achieves an accuracy match with inexpensive computational cost.
暂无评论