Abstract: Visual target tracking is task of locating a target in consecutive frame of a video. Conventional systems observe target behavior frames of the video. However, dealing with this problem is very challenging when video has illumination variations, occlusion, change in size and view of the object due to relative motion between camera and object. Embodiments of the present disclosure addresses this problem by implementing Neural Network (NN), its features and their corresponding gradients. Present disclosure explicitly guides the NN by feeding target object of interest (ToI) defined by a bounding box in the first frame of the video. With this guidance, NN generates target activation map via convolutional features map and their gradient maps, thus giving tentative location of the ToI to further exploit to locate target object precisely by using correlation filter(s) and peak location estimator, thus repeating process for every frame of video to track ToI accurately.

Abstract: Biometric templates (template) are used to store biometric data of one or more users. As biometric data of each user is unique and non-replaceable, template security is a major concern. This disclosure relates generally to template protection, and more particularly to a method and system for biometric template protection. The system generates a final perturbation value for data in the template, based on a first binary code (BC), a second binary code (BC?), feature vectors corresponding to data in the template, and an initial random perturbation. The final perturbation and the feature vectors are mapped to the first binary code. The first binary code represents the template. Every time the first binary code is compromised, the value of first binary code is updated, and the whole process is repeated to generate corresponding final perturbation and then the mapping is done.

Abstract: A classification system according to an embodiment includes a score calculation unit, a determination unit, and a classification unit. The score calculation unit calculates respective scores of predetermined classes from input data. The determination unit determines whether the input data belongs to anyone of the classes based on the respective scores of the classes, which are calculated by the score calculation unit. The classification unit determines which one of the classes the input data belongs to, based on the calculated scores when the determination unit determines that the input data belongs to anyone of the classes and determines that the input data belongs to an unknown class that is other than the classes when the determination unit determines that the input data does not belong the classes.

Abstract: Systems and methods for insertion of a watermark into images and tampering detection of the watermarked images by a Convolutional Neural Network (CNN) technique. The traditional systems and methods provide for detecting the tampering of the watermarked images by simply identifying a presence of an inserted watermark into an image but none them provide for inserting a random sequence into input image(s) and then detect the tampering by classifying the input image(s) by a neural network.

Abstract: Target tracking in a video is a highly challenging problem as the target may be effected by its appearance changes along the video, partial occlusions, background clutter, illumination variations, surrounding environment and also due to changes in the motion of the target. Embodiments of the present disclosure address this problem by implementing neural network for convolution feature maps and their gradient maps generation. The proposed two-class neural network (TCNN) is guided by feeding it target of interest defined by a bounding box in a first frame of the video. With this target guidance TCNN generates target activation map by using convolutional features and gradient maps. Target activation map gives tentative location of target, and this is further exploited to locate target precisely by using correlation filter(s) and peak location estimator based on identified context. This process repeats for every frame of the video to track the target accurately.

Abstract: Biometric templates (template) are used to store biometric data of one or more users. As biometric data of each user is unique and non-replaceable, template security is a major concern. This disclosure relates generally to template protection, and more particularly to a method and system for biometric template protection. The system generates a final perturbation value for data in the template, based on a first binary code (BC), a second binary code (BC?), feature vectors corresponding to data in the template, and an initial random perturbation. The final perturbation and the feature vectors are mapped to the first binary code. The first binary code represents the template. Every time the first binary code is compromised, the value of first binary code is updated, and the whole process is repeated to generate corresponding final perturbation and then the mapping is done.

Abstract: Visual target tracking is task of locating a target in consecutive frame of a video. Conventional systems observe target behavior frames of the video. However, dealing with this problem is very challenging when video has illumination variations, occlusion, change in size and view of the object due to relative motion between camera and object. Embodiments of the present disclosure addresses this problem by implementing Neural Network (NN), its features and their corresponding gradients. Present disclosure explicitly guides the NN by feeding target object of interest (ToI) defined by a bounding box in the first frame of the video. With this guidance, NN generates target activation map via convolutional features map and their gradient maps, thus giving tentative location of the ToI to further exploit to locate target object precisely by using correlation filter(s) and peak location estimator, thus repeating process for every frame of video to track ToI accurately.

Abstract: System and method for text localization in images are disclosed. In an embodiment, a line and graphic eliminated image is received. Further, horizontal projection is performed on rows of the image to obtain a first flag vector, the flag vector indicates whether there is text in each row. Furthermore, a number of run-lengths of consecutive 1's and 0's is computed in the first flag vector. Moreover, text lines is extracted in the image based on the computed number of run-lengths of consecutive 1's and 0's in the first flag vector. Also, vertical projection is performed on the text lines to obtain a second flag vector for the text lines. Further, a number of run-lengths of consecutive 1's and 0's is computed in the second flag vectors. Furthermore, text is localized in the image based on the computed number of run-lengths of consecutive 1's and 0's in the second flag vectors.

Abstract: Target tracking in a video is a highly challenging problem as the target may be effected by its appearance changes along the video, partial occlusions, background clutter, illumination variations, surrounding environment and also due to changes in the motion of the target. Embodiments of the present disclosure address this problem by implementing neural network for convolution feature maps and their gradient maps generation. The proposed two-class neural network (TCNN) is guided by feeding it target of interest defined by a bounding box in a first frame of the video. With this target guidance TCNN generates target activation map by using convolutional features and gradient maps. Target activation map gives tentative location of target, and this is further exploited to locate target precisely by using correlation filter(s) and peak location estimator based on identified context. This process repeats for every frame of the video to track the target accurately.

Abstract: Systems and methods for insertion of a watermark into images and tampering detection of the watermarked images by a Convolutional Neural Network (CNN) technique. The traditional systems and methods provide for detecting the tampering of the watermarked images by simply identifying a presence of an inserted watermark into an image but none them provide for inserting a random sequence into input image(s) and then detect the tampering by classifying the input image(s) by a neural network.

Abstract: System(s) and method(s) for embedding and extracting a watermark from a digital media are described. The method may include embedding a set of watermark bits in a N*N block of the digital media based on a local binary pattern (LBP) synthesis process. The LBP synthesis process comprises changing a value of a neighbor of a central pixel of the N*N block to make the value of the neighbor greater than a value of the central pixel when a watermark bit corresponding to the neighbor is ‘1’ and changing the value of the neighbor of the central pixel to make the value of the neighbor lesser than the value of the central pixel when the watermark bit corresponding to the neighbor is ‘0’.

Abstract: System and method for text localization in images are disclosed. In an embodiment, a line and graphic eliminated image is received. Further, horizontal projection is performed on rows of the image to obtain a first flag vector, the flag vector indicates whether there is text in each row. Furthermore, a number of run-lengths of consecutive 1's and 0's is computed in the first flag vector. Moreover, text lines is extracted in the image based on the computed number of run-lengths of consecutive 1's and 0's in the first flag vector. Also, vertical projection is performed on the text lines to obtain a second flag vector for the text lines. Further, a number of run-lengths of consecutive 1's and 0's is computed in the second flag vectors. Furthermore, text is localized in the image based on the computed number of run-lengths of consecutive 1's and 0's in the second flag vectors.

Abstract: A computer implemented system for audio watermarking for providing robust and blind audio watermarking. The system comprises a watermark embedding system wherein an audio signal is divided into audio frames, multi-level District Wavelet Transform (DWT) is applied on each frame, followed by Singular Value Decomposition (SVD) and embedding the watermark, further followed by inverse SVD and inverse DWT to get watermarked audio frames which are combined to generate a watermarked audio signal. The system further comprises watermark extracting detection system wherein the watermarked audio signal which may be attacked and/or modified is divided into watermarked audio frames, multilevel DWT is applied on each watermarked audio frame, followed by SVD, extracting the embedded watermarked, correlating the extracted watermark with pre-stored watermarks, calculating Peak to Sidelobe ratio (PSR) from the correlation coefficient arrays and finally comparing each PSR with a threshold to authenticate the embedded watermark.

Abstract: A method and system for blind audio watermarking has been envisaged. The system embeds an in-audible watermark by modifying a predetermined singular value in the wavelet domain. The generated watermarked audio signals are robust again Gaussian noise and other watermark removal and compression attacks. Also, the system can perform watermark extraction/detection in real-time without requiring in advance, either the original audio signal or its processed parameters. Thereby, the method and system provide a real-time, efficient technique for blind audio watermarking.

Abstract: A computer implemented system for audio watermarking for providing robust and blind audio watermarking. The system comprises a watermark embedding system wherein an audio signal is divided into audio frames, multi-level District Wavelet Transform (DWT) is applied on each frame, followed by Singular Value Decomposition (SVD) and embedding the watermark, further followed by inverse SVD and inverse DWT to get watermarked audio frames which are combined to generate a watermarked audio signal. The system further comprises watermark extracting detection system wherein the watermarked audio signal which may be attacked and/or modified is divided into watermarked audio frames, multilevel DWT is applied on each watermarked audio frame, followed by SVD, extracting the embedded watermarked, correlating the extracted watermark with pre-stored watermarks, calculating Peak to Sidelobe ratio (PSR) from the correlation coefficient arrays and finally comparing each PSR with a threshold to authenticate the embedded watermark.

Abstract: System(s) and method(s) for embedding and extracting a watermark from a digital media are described. The method may include embedding a set of watermark bits in a N*N block of the digital media based on a local binary pattern (LBP) synthesis process. The LBP synthesis process comprises changing a value of a neighbor of a central pixel of the N*N block to make the value of the neighbor greater than a value of the central pixel when a watermark bit corresponding to the neighbor is ‘1’ and changing the value of the neighbor of the central pixel to make the value of the neighbor lesser than the value of the central pixel when the watermark bit corresponding to the neighbor is ‘0’.

Abstract: The present invention relates to a system and method for routing a data from one or more mobile communication channel to one or more fixed communication channel in a resource efficient manner. The delinking router of the system communicates with the one or more mobile communication channel and with one or more fixed communication channel and transmits the request to a Base Transmitting Station (BTS) for routing the data from the mobile communication channel to the fixed communication channel. The delinking router also reduces energy and spectrum consumption of the mobile communication channel by turning off its radio frequency module after routing its data to the fixed communication channel.

Abstract: System(s) and method(s) for embedding and extracting a watermark from a digital media are described. The method may include generating a plurality of random patterns where the plurality of random patterns includes at least a primary random pattern and a secondary random pattern. The random patterns are indicative of patterns to be embedded to the digital media. The method may further include converting at least the primary random pattern and the secondary random pattern to a bipolar form. The primary random pattern and the secondary random pattern thereof are converted to a primary bipolar random pattern and a secondary bipolar random pattern. The method may also include embedding one or more of the primary bipolar random pattern and the secondary bipolar random pattern in the digital media as the watermark based on a watermark ID, where the watermark ID is a random pattern.

Abstract: Systems and methods for discrete signal synchronization based on a known bit pattern are described. In one aspect of the present subject matter, a discrete signal synchronization system is configured to synchronize a preprocessed discrete signal with a modified discrete signal. The system comprises a processor and a synchronization module coupled to the processor. The synchronization module comprises an extraction module and comparison module. The extraction module determines a bit pattern from the modified discrete signal using Discrete Wavelet Transformation (DWT) and Singular Value Decomposition (SVD). The comparison module compares the determined bit pattern with a known bit pattern of the preprocessed discrete signal and records a time point at which the determined bit pattern matches with the known bit pattern of the preprocessed discrete signal as a synchronization point.

Abstract: Systems and methods for discrete signal synchronization based on a known bit pattern are described. In one aspect of the present subject matter, a discrete signal synchronization system is configured to synchronize a preprocessed discrete signal with a modified discrete signal. The system comprises a processor and a synchronization module coupled to the processor. The synchronization module comprises an extraction module and comparison module. The extraction module determines a bit pattern from the modified discrete signal using Discrete Wavelet Transformation (DWT) and Singular Value Decomposition (SVD). The comparison module compares the determined bit pattern with a known bit pattern of the preprocessed discrete signal and records a time point at which the determined bit pattern matches with the known bit pattern of the preprocessed discrete signal as a synchronization point.